3 * Internationalisation code.
5 * This program is free software; you can redistribute it and/or modify
6 * it under the terms of the GNU General Public License as published by
7 * the Free Software Foundation; either version 2 of the License, or
8 * (at your option) any later version.
10 * This program is distributed in the hope that it will be useful,
11 * but WITHOUT ANY WARRANTY; without even the implied warranty of
12 * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the
13 * GNU General Public License for more details.
15 * You should have received a copy of the GNU General Public License along
16 * with this program; if not, write to the Free Software Foundation, Inc.,
17 * 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301, USA.
18 * http://www.gnu.org/copyleft/gpl.html
25 * @defgroup Language Language
28 if ( !defined( 'MEDIAWIKI' ) ) {
29 echo "This file is part of MediaWiki, it is not a valid entry point.\n";
33 if ( function_exists( 'mb_strtoupper' ) ) {
34 mb_internal_encoding( 'UTF-8' );
38 * Internationalisation code
43 * @var LanguageConverter
47 public $mVariants, $mCode, $mLoaded = false;
48 public $mMagicExtensions = array(), $mMagicHookDone = false;
49 private $mHtmlCode = null, $mParentLanguage = false;
51 public $dateFormatStrings = array();
52 public $mExtendedSpecialPageAliases;
54 protected $namespaceNames, $mNamespaceIds, $namespaceAliases;
57 * ReplacementArray object caches
59 public $transformData = array();
62 * @var LocalisationCache
64 static public $dataCache;
66 static public $mLangObjCache = array();
68 static public $mWeekdayMsgs = array(
69 'sunday', 'monday', 'tuesday', 'wednesday', 'thursday',
73 static public $mWeekdayAbbrevMsgs = array(
74 'sun', 'mon', 'tue', 'wed', 'thu', 'fri', 'sat'
77 static public $mMonthMsgs = array(
78 'january', 'february', 'march', 'april', 'may_long', 'june',
79 'july', 'august', 'september', 'october', 'november',
82 static public $mMonthGenMsgs = array(
83 'january-gen', 'february-gen', 'march-gen', 'april-gen', 'may-gen', 'june-gen',
84 'july-gen', 'august-gen', 'september-gen', 'october-gen', 'november-gen',
87 static public $mMonthAbbrevMsgs = array(
88 'jan', 'feb', 'mar', 'apr', 'may', 'jun', 'jul', 'aug',
89 'sep', 'oct', 'nov', 'dec'
92 static public $mIranianCalendarMonthMsgs = array(
93 'iranian-calendar-m1', 'iranian-calendar-m2', 'iranian-calendar-m3',
94 'iranian-calendar-m4', 'iranian-calendar-m5', 'iranian-calendar-m6',
95 'iranian-calendar-m7', 'iranian-calendar-m8', 'iranian-calendar-m9',
96 'iranian-calendar-m10', 'iranian-calendar-m11', 'iranian-calendar-m12'
99 static public $mHebrewCalendarMonthMsgs = array(
100 'hebrew-calendar-m1', 'hebrew-calendar-m2', 'hebrew-calendar-m3',
101 'hebrew-calendar-m4', 'hebrew-calendar-m5', 'hebrew-calendar-m6',
102 'hebrew-calendar-m7', 'hebrew-calendar-m8', 'hebrew-calendar-m9',
103 'hebrew-calendar-m10', 'hebrew-calendar-m11', 'hebrew-calendar-m12',
104 'hebrew-calendar-m6a', 'hebrew-calendar-m6b'
107 static public $mHebrewCalendarMonthGenMsgs = array(
108 'hebrew-calendar-m1-gen', 'hebrew-calendar-m2-gen', 'hebrew-calendar-m3-gen',
109 'hebrew-calendar-m4-gen', 'hebrew-calendar-m5-gen', 'hebrew-calendar-m6-gen',
110 'hebrew-calendar-m7-gen', 'hebrew-calendar-m8-gen', 'hebrew-calendar-m9-gen',
111 'hebrew-calendar-m10-gen', 'hebrew-calendar-m11-gen', 'hebrew-calendar-m12-gen',
112 'hebrew-calendar-m6a-gen', 'hebrew-calendar-m6b-gen'
115 static public $mHijriCalendarMonthMsgs = array(
116 'hijri-calendar-m1', 'hijri-calendar-m2', 'hijri-calendar-m3',
117 'hijri-calendar-m4', 'hijri-calendar-m5', 'hijri-calendar-m6',
118 'hijri-calendar-m7', 'hijri-calendar-m8', 'hijri-calendar-m9',
119 'hijri-calendar-m10', 'hijri-calendar-m11', 'hijri-calendar-m12'
126 static public $durationIntervals = array(
127 'millennia' => 31556952000,
128 'centuries' => 3155695200,
129 'decades' => 315569520,
130 'years' => 31556952, // 86400 * ( 365 + ( 24 * 3 + 25 ) / 400 )
139 * Cache for language fallbacks.
140 * @see Language::getFallbacksIncludingSiteLanguage
144 static private $fallbackLanguageCache = array();
147 * Cache for language names
148 * @var MapCacheLRU|null
150 static private $languageNameCache;
153 * Get a cached or new language object for a given language code
154 * @param string $code
157 static function factory( $code ) {
158 global $wgDummyLanguageCodes, $wgLangObjCacheSize;
160 if ( isset( $wgDummyLanguageCodes[$code] ) ) {
161 $code = $wgDummyLanguageCodes[$code];
164 // get the language object to process
165 $langObj = isset( self
::$mLangObjCache[$code] )
166 ? self
::$mLangObjCache[$code]
167 : self
::newFromCode( $code );
169 // merge the language object in to get it up front in the cache
170 self
::$mLangObjCache = array_merge( array( $code => $langObj ), self
::$mLangObjCache );
171 // get rid of the oldest ones in case we have an overflow
172 self
::$mLangObjCache = array_slice( self
::$mLangObjCache, 0, $wgLangObjCacheSize, true );
178 * Create a language object for a given language code
179 * @param string $code
180 * @throws MWException
183 protected static function newFromCode( $code ) {
184 // Protect against path traversal below
185 if ( !Language
::isValidCode( $code )
186 ||
strcspn( $code, ":/\\\000" ) !== strlen( $code )
188 throw new MWException( "Invalid language code \"$code\"" );
191 if ( !Language
::isValidBuiltInCode( $code ) ) {
192 // It's not possible to customise this code with class files, so
193 // just return a Language object. This is to support uselang= hacks.
194 $lang = new Language
;
195 $lang->setCode( $code );
199 // Check if there is a language class for the code
200 $class = self
::classFromCode( $code );
201 self
::preloadLanguageClass( $class );
202 if ( class_exists( $class ) ) {
207 // Keep trying the fallback list until we find an existing class
208 $fallbacks = Language
::getFallbacksFor( $code );
209 foreach ( $fallbacks as $fallbackCode ) {
210 if ( !Language
::isValidBuiltInCode( $fallbackCode ) ) {
211 throw new MWException( "Invalid fallback '$fallbackCode' in fallback sequence for '$code'" );
214 $class = self
::classFromCode( $fallbackCode );
215 self
::preloadLanguageClass( $class );
216 if ( class_exists( $class ) ) {
217 $lang = Language
::newFromCode( $fallbackCode );
218 $lang->setCode( $code );
223 throw new MWException( "Invalid fallback sequence for language '$code'" );
227 * Checks whether any localisation is available for that language tag
228 * in MediaWiki (MessagesXx.php exists).
230 * @param string $code Language tag (in lower case)
231 * @return bool Whether language is supported
234 public static function isSupportedLanguage( $code ) {
235 return self
::isValidBuiltInCode( $code )
236 && ( is_readable( self
::getMessagesFileName( $code ) )
237 ||
is_readable( self
::getJsonMessagesFileName( $code ) )
242 * Returns true if a language code string is a well-formed language tag
243 * according to RFC 5646.
244 * This function only checks well-formedness; it doesn't check that
245 * language, script or variant codes actually exist in the repositories.
247 * Based on regexes by Mark Davis of the Unicode Consortium:
248 * http://unicode.org/repos/cldr/trunk/tools/java/org/unicode/cldr/util/data/langtagRegex.txt
250 * @param string $code
251 * @param bool $lenient Whether to allow '_' as separator. The default is only '-'.
256 public static function isWellFormedLanguageTag( $code, $lenient = false ) {
259 $alphanum = '[a-z0-9]';
260 $x = 'x'; # private use singleton
261 $singleton = '[a-wy-z]'; # other singleton
262 $s = $lenient ?
'[-_]' : '-';
264 $language = "$alpha{2,8}|$alpha{2,3}$s$alpha{3}";
265 $script = "$alpha{4}"; # ISO 15924
266 $region = "(?:$alpha{2}|$digit{3})"; # ISO 3166-1 alpha-2 or UN M.49
267 $variant = "(?:$alphanum{5,8}|$digit$alphanum{3})";
268 $extension = "$singleton(?:$s$alphanum{2,8})+";
269 $privateUse = "$x(?:$s$alphanum{1,8})+";
271 # Define certain grandfathered codes, since otherwise the regex is pretty useless.
272 # Since these are limited, this is safe even later changes to the registry --
273 # the only oddity is that it might change the type of the tag, and thus
274 # the results from the capturing groups.
275 # http://www.iana.org/assignments/language-subtag-registry
277 $grandfathered = "en{$s}GB{$s}oed"
278 . "|i{$s}(?:ami|bnn|default|enochian|hak|klingon|lux|mingo|navajo|pwn|tao|tay|tsu)"
279 . "|no{$s}(?:bok|nyn)"
280 . "|sgn{$s}(?:BE{$s}(?:fr|nl)|CH{$s}de)"
281 . "|zh{$s}min{$s}nan";
283 $variantList = "$variant(?:$s$variant)*";
284 $extensionList = "$extension(?:$s$extension)*";
286 $langtag = "(?:($language)"
289 . "(?:$s$variantList)?"
290 . "(?:$s$extensionList)?"
291 . "(?:$s$privateUse)?)";
293 # The final breakdown, with capturing groups for each of these components
294 # The variants, extensions, grandfathered, and private-use may have interior '-'
296 $root = "^(?:$langtag|$privateUse|$grandfathered)$";
298 return (bool)preg_match( "/$root/", strtolower( $code ) );
302 * Returns true if a language code string is of a valid form, whether or
303 * not it exists. This includes codes which are used solely for
304 * customisation via the MediaWiki namespace.
306 * @param string $code
310 public static function isValidCode( $code ) {
311 static $cache = array();
312 if ( isset( $cache[$code] ) ) {
313 return $cache[$code];
315 // People think language codes are html safe, so enforce it.
316 // Ideally we should only allow a-zA-Z0-9-
317 // but, .+ and other chars are often used for {{int:}} hacks
318 // see bugs 37564, 37587, 36938
320 strcspn( $code, ":/\\\000&<>'\"" ) === strlen( $code )
321 && !preg_match( MediaWikiTitleCodec
::getTitleInvalidRegex(), $code );
323 return $cache[$code];
327 * Returns true if a language code is of a valid form for the purposes of
328 * internal customisation of MediaWiki, via Messages*.php or *.json.
330 * @param string $code
332 * @throws MWException
336 public static function isValidBuiltInCode( $code ) {
338 if ( !is_string( $code ) ) {
339 if ( is_object( $code ) ) {
340 $addmsg = " of class " . get_class( $code );
344 $type = gettype( $code );
345 throw new MWException( __METHOD__
. " must be passed a string, $type given$addmsg" );
348 return (bool)preg_match( '/^[a-z0-9-]{2,}$/', $code );
352 * Returns true if a language code is an IETF tag known to MediaWiki.
359 public static function isKnownLanguageTag( $tag ) {
360 static $coreLanguageNames;
362 // Quick escape for invalid input to avoid exceptions down the line
363 // when code tries to process tags which are not valid at all.
364 if ( !self
::isValidBuiltInCode( $tag ) ) {
368 if ( $coreLanguageNames === null ) {
370 include "$IP/languages/Names.php";
373 if ( isset( $coreLanguageNames[$tag] )
374 || self
::fetchLanguageName( $tag, $tag ) !== ''
383 * @param string $code
384 * @return string Name of the language class
386 public static function classFromCode( $code ) {
387 if ( $code == 'en' ) {
390 return 'Language' . str_replace( '-', '_', ucfirst( $code ) );
395 * Includes language class files
397 * @param string $class Name of the language class
399 public static function preloadLanguageClass( $class ) {
402 if ( $class === 'Language' ) {
406 if ( file_exists( "$IP/languages/classes/$class.php" ) ) {
407 include_once "$IP/languages/classes/$class.php";
412 * Get the LocalisationCache instance
414 * @return LocalisationCache
416 public static function getLocalisationCache() {
417 if ( is_null( self
::$dataCache ) ) {
418 global $wgLocalisationCacheConf;
419 $class = $wgLocalisationCacheConf['class'];
420 self
::$dataCache = new $class( $wgLocalisationCacheConf );
422 return self
::$dataCache;
425 function __construct() {
426 $this->mConverter
= new FakeConverter( $this );
427 // Set the code to the name of the descendant
428 if ( get_class( $this ) == 'Language' ) {
431 $this->mCode
= str_replace( '_', '-', strtolower( substr( get_class( $this ), 8 ) ) );
433 self
::getLocalisationCache();
437 * Reduce memory usage
439 function __destruct() {
440 foreach ( $this as $name => $value ) {
441 unset( $this->$name );
446 * Hook which will be called if this is the content language.
447 * Descendants can use this to register hook functions or modify globals
449 function initContLang() {
456 function getFallbackLanguages() {
457 return self
::getFallbacksFor( $this->mCode
);
461 * Exports $wgBookstoreListEn
464 function getBookstoreList() {
465 return self
::$dataCache->getItem( $this->mCode
, 'bookstoreList' );
469 * Returns an array of localised namespaces indexed by their numbers. If the namespace is not
470 * available in localised form, it will be included in English.
474 public function getNamespaces() {
475 if ( is_null( $this->namespaceNames
) ) {
476 global $wgMetaNamespace, $wgMetaNamespaceTalk, $wgExtraNamespaces;
478 $this->namespaceNames
= self
::$dataCache->getItem( $this->mCode
, 'namespaceNames' );
479 $validNamespaces = MWNamespace
::getCanonicalNamespaces();
481 $this->namespaceNames
= $wgExtraNamespaces +
$this->namespaceNames +
$validNamespaces;
483 $this->namespaceNames
[NS_PROJECT
] = $wgMetaNamespace;
484 if ( $wgMetaNamespaceTalk ) {
485 $this->namespaceNames
[NS_PROJECT_TALK
] = $wgMetaNamespaceTalk;
487 $talk = $this->namespaceNames
[NS_PROJECT_TALK
];
488 $this->namespaceNames
[NS_PROJECT_TALK
] =
489 $this->fixVariableInNamespace( $talk );
492 # Sometimes a language will be localised but not actually exist on this wiki.
493 foreach ( $this->namespaceNames
as $key => $text ) {
494 if ( !isset( $validNamespaces[$key] ) ) {
495 unset( $this->namespaceNames
[$key] );
499 # The above mixing may leave namespaces out of canonical order.
500 # Re-order by namespace ID number...
501 ksort( $this->namespaceNames
);
503 Hooks
::run( 'LanguageGetNamespaces', array( &$this->namespaceNames
) );
506 return $this->namespaceNames
;
510 * Arbitrarily set all of the namespace names at once. Mainly used for testing
511 * @param array $namespaces Array of namespaces (id => name)
513 public function setNamespaces( array $namespaces ) {
514 $this->namespaceNames
= $namespaces;
515 $this->mNamespaceIds
= null;
519 * Resets all of the namespace caches. Mainly used for testing
521 public function resetNamespaces() {
522 $this->namespaceNames
= null;
523 $this->mNamespaceIds
= null;
524 $this->namespaceAliases
= null;
528 * A convenience function that returns the same thing as
529 * getNamespaces() except with the array values changed to ' '
530 * where it found '_', useful for producing output to be displayed
531 * e.g. in <select> forms.
535 function getFormattedNamespaces() {
536 $ns = $this->getNamespaces();
537 foreach ( $ns as $k => $v ) {
538 $ns[$k] = strtr( $v, '_', ' ' );
544 * Get a namespace value by key
546 * $mw_ns = $wgContLang->getNsText( NS_MEDIAWIKI );
547 * echo $mw_ns; // prints 'MediaWiki'
550 * @param int $index The array key of the namespace to return
551 * @return string|bool String if the namespace value exists, otherwise false
553 function getNsText( $index ) {
554 $ns = $this->getNamespaces();
556 return isset( $ns[$index] ) ?
$ns[$index] : false;
560 * A convenience function that returns the same thing as
561 * getNsText() except with '_' changed to ' ', useful for
565 * $mw_ns = $wgContLang->getFormattedNsText( NS_MEDIAWIKI_TALK );
566 * echo $mw_ns; // prints 'MediaWiki talk'
569 * @param int $index The array key of the namespace to return
570 * @return string Namespace name without underscores (empty string if namespace does not exist)
572 function getFormattedNsText( $index ) {
573 $ns = $this->getNsText( $index );
575 return strtr( $ns, '_', ' ' );
579 * Returns gender-dependent namespace alias if available.
580 * See https://www.mediawiki.org/wiki/Manual:$wgExtraGenderNamespaces
581 * @param int $index Namespace index
582 * @param string $gender Gender key (male, female... )
586 function getGenderNsText( $index, $gender ) {
587 global $wgExtraGenderNamespaces;
589 $ns = $wgExtraGenderNamespaces +
590 (array)self
::$dataCache->getItem( $this->mCode
, 'namespaceGenderAliases' );
592 return isset( $ns[$index][$gender] ) ?
$ns[$index][$gender] : $this->getNsText( $index );
596 * Whether this language uses gender-dependent namespace aliases.
597 * See https://www.mediawiki.org/wiki/Manual:$wgExtraGenderNamespaces
601 function needsGenderDistinction() {
602 global $wgExtraGenderNamespaces, $wgExtraNamespaces;
603 if ( count( $wgExtraGenderNamespaces ) > 0 ) {
604 // $wgExtraGenderNamespaces overrides everything
606 } elseif ( isset( $wgExtraNamespaces[NS_USER
] ) && isset( $wgExtraNamespaces[NS_USER_TALK
] ) ) {
607 /// @todo There may be other gender namespace than NS_USER & NS_USER_TALK in the future
608 // $wgExtraNamespaces overrides any gender aliases specified in i18n files
611 // Check what is in i18n files
612 $aliases = self
::$dataCache->getItem( $this->mCode
, 'namespaceGenderAliases' );
613 return count( $aliases ) > 0;
618 * Get a namespace key by value, case insensitive.
619 * Only matches namespace names for the current language, not the
620 * canonical ones defined in Namespace.php.
622 * @param string $text
623 * @return int|bool An integer if $text is a valid value otherwise false
625 function getLocalNsIndex( $text ) {
626 $lctext = $this->lc( $text );
627 $ids = $this->getNamespaceIds();
628 return isset( $ids[$lctext] ) ?
$ids[$lctext] : false;
634 function getNamespaceAliases() {
635 if ( is_null( $this->namespaceAliases
) ) {
636 $aliases = self
::$dataCache->getItem( $this->mCode
, 'namespaceAliases' );
640 foreach ( $aliases as $name => $index ) {
641 if ( $index === NS_PROJECT_TALK
) {
642 unset( $aliases[$name] );
643 $name = $this->fixVariableInNamespace( $name );
644 $aliases[$name] = $index;
649 global $wgExtraGenderNamespaces;
650 $genders = $wgExtraGenderNamespaces +
651 (array)self
::$dataCache->getItem( $this->mCode
, 'namespaceGenderAliases' );
652 foreach ( $genders as $index => $forms ) {
653 foreach ( $forms as $alias ) {
654 $aliases[$alias] = $index;
658 # Also add converted namespace names as aliases, to avoid confusion.
659 $convertedNames = array();
660 foreach ( $this->getVariants() as $variant ) {
661 if ( $variant === $this->mCode
) {
664 foreach ( $this->getNamespaces() as $ns => $_ ) {
665 $convertedNames[$this->getConverter()->convertNamespace( $ns, $variant )] = $ns;
669 $this->namespaceAliases
= $aliases +
$convertedNames;
672 return $this->namespaceAliases
;
678 function getNamespaceIds() {
679 if ( is_null( $this->mNamespaceIds
) ) {
680 global $wgNamespaceAliases;
681 # Put namespace names and aliases into a hashtable.
682 # If this is too slow, then we should arrange it so that it is done
683 # before caching. The catch is that at pre-cache time, the above
684 # class-specific fixup hasn't been done.
685 $this->mNamespaceIds
= array();
686 foreach ( $this->getNamespaces() as $index => $name ) {
687 $this->mNamespaceIds
[$this->lc( $name )] = $index;
689 foreach ( $this->getNamespaceAliases() as $name => $index ) {
690 $this->mNamespaceIds
[$this->lc( $name )] = $index;
692 if ( $wgNamespaceAliases ) {
693 foreach ( $wgNamespaceAliases as $name => $index ) {
694 $this->mNamespaceIds
[$this->lc( $name )] = $index;
698 return $this->mNamespaceIds
;
702 * Get a namespace key by value, case insensitive. Canonical namespace
703 * names override custom ones defined for the current language.
705 * @param string $text
706 * @return int|bool An integer if $text is a valid value otherwise false
708 function getNsIndex( $text ) {
709 $lctext = $this->lc( $text );
710 $ns = MWNamespace
::getCanonicalIndex( $lctext );
711 if ( $ns !== null ) {
714 $ids = $this->getNamespaceIds();
715 return isset( $ids[$lctext] ) ?
$ids[$lctext] : false;
719 * short names for language variants used for language conversion links.
721 * @param string $code
722 * @param bool $usemsg Use the "variantname-xyz" message if it exists
725 function getVariantname( $code, $usemsg = true ) {
726 $msg = "variantname-$code";
727 if ( $usemsg && wfMessage( $msg )->exists() ) {
728 return $this->getMessageFromDB( $msg );
730 $name = self
::fetchLanguageName( $code );
732 return $name; # if it's defined as a language name, show that
734 # otherwise, output the language code
740 * @deprecated since 1.24, doesn't handle conflicting aliases. Use
741 * SpecialPageFactory::getLocalNameFor instead.
742 * @param string $name
745 function specialPage( $name ) {
746 $aliases = $this->getSpecialPageAliases();
747 if ( isset( $aliases[$name][0] ) ) {
748 $name = $aliases[$name][0];
750 return $this->getNsText( NS_SPECIAL
) . ':' . $name;
756 function getDatePreferences() {
757 return self
::$dataCache->getItem( $this->mCode
, 'datePreferences' );
763 function getDateFormats() {
764 return self
::$dataCache->getItem( $this->mCode
, 'dateFormats' );
768 * @return array|string
770 function getDefaultDateFormat() {
771 $df = self
::$dataCache->getItem( $this->mCode
, 'defaultDateFormat' );
772 if ( $df === 'dmy or mdy' ) {
773 global $wgAmericanDates;
774 return $wgAmericanDates ?
'mdy' : 'dmy';
783 function getDatePreferenceMigrationMap() {
784 return self
::$dataCache->getItem( $this->mCode
, 'datePreferenceMigrationMap' );
788 * @param string $image
791 function getImageFile( $image ) {
792 return self
::$dataCache->getSubitem( $this->mCode
, 'imageFiles', $image );
799 function getImageFiles() {
800 return self
::$dataCache->getItem( $this->mCode
, 'imageFiles' );
806 function getExtraUserToggles() {
807 return (array)self
::$dataCache->getItem( $this->mCode
, 'extraUserToggles' );
814 function getUserToggle( $tog ) {
815 return $this->getMessageFromDB( "tog-$tog" );
819 * Get native language names, indexed by code.
820 * Only those defined in MediaWiki, no other data like CLDR.
821 * If $customisedOnly is true, only returns codes with a messages file
823 * @param bool $customisedOnly
826 * @deprecated since 1.20, use fetchLanguageNames()
828 public static function getLanguageNames( $customisedOnly = false ) {
829 return self
::fetchLanguageNames( null, $customisedOnly ?
'mwfile' : 'mw' );
833 * Get translated language names. This is done on best effort and
834 * by default this is exactly the same as Language::getLanguageNames.
835 * The CLDR extension provides translated names.
836 * @param string $code Language code.
837 * @return array Language code => language name
839 * @deprecated since 1.20, use fetchLanguageNames()
841 public static function getTranslatedLanguageNames( $code ) {
842 return self
::fetchLanguageNames( $code, 'all' );
846 * Get an array of language names, indexed by code.
847 * @param null|string $inLanguage Code of language in which to return the names
848 * Use null for autonyms (native names)
849 * @param string $include One of:
850 * 'all' all available languages
851 * 'mw' only if the language is defined in MediaWiki or wgExtraLanguageNames (default)
852 * 'mwfile' only if the language is in 'mw' *and* has a message file
853 * @return array Language code => language name
856 public static function fetchLanguageNames( $inLanguage = null, $include = 'mw' ) {
857 $cacheKey = $inLanguage === null ?
'null' : $inLanguage;
858 $cacheKey .= ":$include";
859 if ( self
::$languageNameCache === null ) {
860 self
::$languageNameCache = new MapCacheLRU( 20 );
862 if ( self
::$languageNameCache->has( $cacheKey ) ) {
863 $ret = self
::$languageNameCache->get( $cacheKey );
865 $ret = self
::fetchLanguageNamesUncached( $inLanguage, $include );
866 self
::$languageNameCache->set( $cacheKey, $ret );
872 * Uncached helper for fetchLanguageNames
873 * @param null|string $inLanguage Code of language in which to return the names
874 * Use null for autonyms (native names)
875 * @param string $include One of:
876 * 'all' all available languages
877 * 'mw' only if the language is defined in MediaWiki or wgExtraLanguageNames (default)
878 * 'mwfile' only if the language is in 'mw' *and* has a message file
879 * @return array Language code => language name
881 private static function fetchLanguageNamesUncached( $inLanguage = null, $include = 'mw' ) {
882 global $wgExtraLanguageNames;
883 static $coreLanguageNames;
885 if ( $coreLanguageNames === null ) {
887 include "$IP/languages/Names.php";
890 // If passed an invalid language code to use, fallback to en
891 if ( $inLanguage !== null && !Language
::isValidCode( $inLanguage ) ) {
898 # TODO: also include when $inLanguage is null, when this code is more efficient
899 Hooks
::run( 'LanguageGetTranslatedLanguageNames', array( &$names, $inLanguage ) );
902 $mwNames = $wgExtraLanguageNames +
$coreLanguageNames;
903 foreach ( $mwNames as $mwCode => $mwName ) {
904 # - Prefer own MediaWiki native name when not using the hook
905 # - For other names just add if not added through the hook
906 if ( $mwCode === $inLanguage ||
!isset( $names[$mwCode] ) ) {
907 $names[$mwCode] = $mwName;
911 if ( $include === 'all' ) {
917 $coreCodes = array_keys( $mwNames );
918 foreach ( $coreCodes as $coreCode ) {
919 $returnMw[$coreCode] = $names[$coreCode];
922 if ( $include === 'mwfile' ) {
923 $namesMwFile = array();
924 # We do this using a foreach over the codes instead of a directory
925 # loop so that messages files in extensions will work correctly.
926 foreach ( $returnMw as $code => $value ) {
927 if ( is_readable( self
::getMessagesFileName( $code ) )
928 ||
is_readable( self
::getJsonMessagesFileName( $code ) )
930 $namesMwFile[$code] = $names[$code];
934 ksort( $namesMwFile );
939 # 'mw' option; default if it's not one of the other two options (all/mwfile)
944 * @param string $code The code of the language for which to get the name
945 * @param null|string $inLanguage Code of language in which to return the name (null for autonyms)
946 * @param string $include 'all', 'mw' or 'mwfile'; see fetchLanguageNames()
947 * @return string Language name or empty
950 public static function fetchLanguageName( $code, $inLanguage = null, $include = 'all' ) {
951 $code = strtolower( $code );
952 $array = self
::fetchLanguageNames( $inLanguage, $include );
953 return !array_key_exists( $code, $array ) ?
'' : $array[$code];
957 * Get a message from the MediaWiki namespace.
959 * @param string $msg Message name
962 function getMessageFromDB( $msg ) {
963 return $this->msg( $msg )->text();
967 * Get message object in this language. Only for use inside this class.
969 * @param string $msg Message name
972 protected function msg( $msg ) {
973 return wfMessage( $msg )->inLanguage( $this );
977 * Get the native language name of $code.
978 * Only if defined in MediaWiki, no other data like CLDR.
979 * @param string $code
981 * @deprecated since 1.20, use fetchLanguageName()
983 function getLanguageName( $code ) {
984 return self
::fetchLanguageName( $code );
991 function getMonthName( $key ) {
992 return $this->getMessageFromDB( self
::$mMonthMsgs[$key - 1] );
998 function getMonthNamesArray() {
999 $monthNames = array( '' );
1000 for ( $i = 1; $i < 13; $i++
) {
1001 $monthNames[] = $this->getMonthName( $i );
1007 * @param string $key
1010 function getMonthNameGen( $key ) {
1011 return $this->getMessageFromDB( self
::$mMonthGenMsgs[$key - 1] );
1015 * @param string $key
1018 function getMonthAbbreviation( $key ) {
1019 return $this->getMessageFromDB( self
::$mMonthAbbrevMsgs[$key - 1] );
1025 function getMonthAbbreviationsArray() {
1026 $monthNames = array( '' );
1027 for ( $i = 1; $i < 13; $i++
) {
1028 $monthNames[] = $this->getMonthAbbreviation( $i );
1034 * @param string $key
1037 function getWeekdayName( $key ) {
1038 return $this->getMessageFromDB( self
::$mWeekdayMsgs[$key - 1] );
1042 * @param string $key
1045 function getWeekdayAbbreviation( $key ) {
1046 return $this->getMessageFromDB( self
::$mWeekdayAbbrevMsgs[$key - 1] );
1050 * @param string $key
1053 function getIranianCalendarMonthName( $key ) {
1054 return $this->getMessageFromDB( self
::$mIranianCalendarMonthMsgs[$key - 1] );
1058 * @param string $key
1061 function getHebrewCalendarMonthName( $key ) {
1062 return $this->getMessageFromDB( self
::$mHebrewCalendarMonthMsgs[$key - 1] );
1066 * @param string $key
1069 function getHebrewCalendarMonthNameGen( $key ) {
1070 return $this->getMessageFromDB( self
::$mHebrewCalendarMonthGenMsgs[$key - 1] );
1074 * @param string $key
1077 function getHijriCalendarMonthName( $key ) {
1078 return $this->getMessageFromDB( self
::$mHijriCalendarMonthMsgs[$key - 1] );
1082 * Pass through result from $dateTimeObj->format()
1083 * @param DateTime|bool|null &$dateTimeObj
1085 * @param DateTimeZone|bool|null $zone
1086 * @param string $code
1089 private static function dateTimeObjFormat( &$dateTimeObj, $ts, $zone, $code ) {
1090 if ( !$dateTimeObj ) {
1091 $dateTimeObj = DateTime
::createFromFormat(
1092 'YmdHis', $ts, $zone ?
: new DateTimeZone( 'UTC' )
1095 return $dateTimeObj->format( $code );
1099 * This is a workalike of PHP's date() function, but with better
1100 * internationalisation, a reduced set of format characters, and a better
1103 * Supported format characters are dDjlNwzWFmMntLoYyaAgGhHiscrUeIOPTZ. See
1104 * the PHP manual for definitions. There are a number of extensions, which
1107 * xn Do not translate digits of the next numeric format character
1108 * xN Toggle raw digit (xn) flag, stays set until explicitly unset
1109 * xr Use roman numerals for the next numeric format character
1110 * xh Use hebrew numerals for the next numeric format character
1112 * xg Genitive month name
1114 * xij j (day number) in Iranian calendar
1115 * xiF F (month name) in Iranian calendar
1116 * xin n (month number) in Iranian calendar
1117 * xiy y (two digit year) in Iranian calendar
1118 * xiY Y (full year) in Iranian calendar
1120 * xjj j (day number) in Hebrew calendar
1121 * xjF F (month name) in Hebrew calendar
1122 * xjt t (days in month) in Hebrew calendar
1123 * xjx xg (genitive month name) in Hebrew calendar
1124 * xjn n (month number) in Hebrew calendar
1125 * xjY Y (full year) in Hebrew calendar
1127 * xmj j (day number) in Hijri calendar
1128 * xmF F (month name) in Hijri calendar
1129 * xmn n (month number) in Hijri calendar
1130 * xmY Y (full year) in Hijri calendar
1132 * xkY Y (full year) in Thai solar calendar. Months and days are
1133 * identical to the Gregorian calendar
1134 * xoY Y (full year) in Minguo calendar or Juche year.
1135 * Months and days are identical to the
1136 * Gregorian calendar
1137 * xtY Y (full year) in Japanese nengo. Months and days are
1138 * identical to the Gregorian calendar
1140 * Characters enclosed in double quotes will be considered literal (with
1141 * the quotes themselves removed). Unmatched quotes will be considered
1142 * literal quotes. Example:
1144 * "The month is" F => The month is January
1147 * Backslash escaping is also supported.
1149 * Input timestamp is assumed to be pre-normalized to the desired local
1150 * time zone, if any. Note that the format characters crUeIOPTZ will assume
1151 * $ts is UTC if $zone is not given.
1153 * @param string $format
1154 * @param string $ts 14-character timestamp
1157 * @param DateTimeZone $zone Timezone of $ts
1158 * @param[out] int $ttl The amount of time (in seconds) the output may be cached for.
1159 * Only makes sense if $ts is the current time.
1160 * @todo handling of "o" format character for Iranian, Hebrew, Hijri & Thai?
1162 * @throws MWException
1165 function sprintfDate( $format, $ts, DateTimeZone
$zone = null, &$ttl = null ) {
1170 $dateTimeObj = false;
1179 $usedSecond = false;
1180 $usedMinute = false;
1187 $usedISOYear = false;
1188 $usedIsLeapYear = false;
1190 $usedHebrewMonth = false;
1191 $usedIranianMonth = false;
1192 $usedHijriMonth = false;
1193 $usedHebrewYear = false;
1194 $usedIranianYear = false;
1195 $usedHijriYear = false;
1196 $usedTennoYear = false;
1198 if ( strlen( $ts ) !== 14 ) {
1199 throw new MWException( __METHOD__
. ": The timestamp $ts should have 14 characters" );
1202 if ( !ctype_digit( $ts ) ) {
1203 throw new MWException( __METHOD__
. ": The timestamp $ts should be a number" );
1206 $formatLength = strlen( $format );
1207 for ( $p = 0; $p < $formatLength; $p++
) {
1209 $code = $format[$p];
1210 if ( $code == 'x' && $p < $formatLength - 1 ) {
1211 $code .= $format[++
$p];
1214 if ( ( $code === 'xi'
1220 && $p < $formatLength - 1 ) {
1221 $code .= $format[++
$p];
1232 $rawToggle = !$rawToggle;
1242 $s .= $this->getMonthNameGen( substr( $ts, 4, 2 ) );
1245 $usedHebrewMonth = true;
1247 $hebrew = self
::tsToHebrew( $ts );
1249 $s .= $this->getHebrewCalendarMonthNameGen( $hebrew[1] );
1253 $num = substr( $ts, 6, 2 );
1257 $s .= $this->getWeekdayAbbreviation(
1258 Language
::dateTimeObjFormat( $dateTimeObj, $ts, $zone, 'w' ) +
1
1263 $num = intval( substr( $ts, 6, 2 ) );
1268 $iranian = self
::tsToIranian( $ts );
1275 $hijri = self
::tsToHijri( $ts );
1282 $hebrew = self
::tsToHebrew( $ts );
1288 $s .= $this->getWeekdayName(
1289 Language
::dateTimeObjFormat( $dateTimeObj, $ts, $zone, 'w' ) +
1
1294 $s .= $this->getMonthName( substr( $ts, 4, 2 ) );
1297 $usedIranianMonth = true;
1299 $iranian = self
::tsToIranian( $ts );
1301 $s .= $this->getIranianCalendarMonthName( $iranian[1] );
1304 $usedHijriMonth = true;
1306 $hijri = self
::tsToHijri( $ts );
1308 $s .= $this->getHijriCalendarMonthName( $hijri[1] );
1311 $usedHebrewMonth = true;
1313 $hebrew = self
::tsToHebrew( $ts );
1315 $s .= $this->getHebrewCalendarMonthName( $hebrew[1] );
1319 $num = substr( $ts, 4, 2 );
1323 $s .= $this->getMonthAbbreviation( substr( $ts, 4, 2 ) );
1327 $num = intval( substr( $ts, 4, 2 ) );
1330 $usedIranianMonth = true;
1332 $iranian = self
::tsToIranian( $ts );
1337 $usedHijriMonth = true;
1339 $hijri = self
::tsToHijri ( $ts );
1344 $usedHebrewMonth = true;
1346 $hebrew = self
::tsToHebrew( $ts );
1351 $usedHebrewMonth = true;
1353 $hebrew = self
::tsToHebrew( $ts );
1359 $num = substr( $ts, 0, 4 );
1362 $usedIranianYear = true;
1364 $iranian = self
::tsToIranian( $ts );
1369 $usedHijriYear = true;
1371 $hijri = self
::tsToHijri( $ts );
1376 $usedHebrewYear = true;
1378 $hebrew = self
::tsToHebrew( $ts );
1385 $thai = self
::tsToYear( $ts, 'thai' );
1392 $minguo = self
::tsToYear( $ts, 'minguo' );
1397 $usedTennoYear = true;
1399 $tenno = self
::tsToYear( $ts, 'tenno' );
1405 $num = substr( $ts, 2, 2 );
1408 $usedIranianYear = true;
1410 $iranian = self
::tsToIranian( $ts );
1412 $num = substr( $iranian[0], -2 );
1416 $s .= intval( substr( $ts, 8, 2 ) ) < 12 ?
'am' : 'pm';
1420 $s .= intval( substr( $ts, 8, 2 ) ) < 12 ?
'AM' : 'PM';
1424 $h = substr( $ts, 8, 2 );
1425 $num = $h %
12 ?
$h %
12 : 12;
1429 $num = intval( substr( $ts, 8, 2 ) );
1433 $h = substr( $ts, 8, 2 );
1434 $num = sprintf( '%02d', $h %
12 ?
$h %
12 : 12 );
1438 $num = substr( $ts, 8, 2 );
1442 $num = substr( $ts, 10, 2 );
1446 $num = substr( $ts, 12, 2 );
1456 $s .= Language
::dateTimeObjFormat( $dateTimeObj, $ts, $zone, $code );
1462 $num = Language
::dateTimeObjFormat( $dateTimeObj, $ts, $zone, $code );
1466 $num = Language
::dateTimeObjFormat( $dateTimeObj, $ts, $zone, $code );
1470 $num = Language
::dateTimeObjFormat( $dateTimeObj, $ts, $zone, $code );
1473 $usedIsLeapYear = true;
1474 $num = Language
::dateTimeObjFormat( $dateTimeObj, $ts, $zone, $code );
1477 $usedISOYear = true;
1478 $num = Language
::dateTimeObjFormat( $dateTimeObj, $ts, $zone, $code );
1485 $num = Language
::dateTimeObjFormat( $dateTimeObj, $ts, $zone, $code );
1488 # Backslash escaping
1489 if ( $p < $formatLength - 1 ) {
1490 $s .= $format[++
$p];
1497 if ( $p < $formatLength - 1 ) {
1498 $endQuote = strpos( $format, '"', $p +
1 );
1499 if ( $endQuote === false ) {
1500 # No terminating quote, assume literal "
1503 $s .= substr( $format, $p +
1, $endQuote - $p - 1 );
1507 # Quote at end of string, assume literal "
1514 if ( $num !== false ) {
1515 if ( $rawToggle ||
$raw ) {
1518 } elseif ( $roman ) {
1519 $s .= Language
::romanNumeral( $num );
1521 } elseif ( $hebrewNum ) {
1522 $s .= self
::hebrewNumeral( $num );
1525 $s .= $this->formatNum( $num, true );
1530 if ( $usedSecond ) {
1532 } elseif ( $usedMinute ) {
1533 $ttl = 60 - substr( $ts, 12, 2 );
1534 } elseif ( $usedHour ) {
1535 $ttl = 3600 - substr( $ts, 10, 2 ) * 60 - substr( $ts, 12, 2 );
1536 } elseif ( $usedAMPM ) {
1537 $ttl = 43200 - ( substr( $ts, 8, 2 ) %
12 ) * 3600 -
1538 substr( $ts, 10, 2 ) * 60 - substr( $ts, 12, 2 );
1542 $usedIranianMonth ||
1549 // @todo Someone who understands the non-Gregorian calendars
1550 // should write proper logic for them so that they don't need purged every day.
1551 $ttl = 86400 - substr( $ts, 8, 2 ) * 3600 -
1552 substr( $ts, 10, 2 ) * 60 - substr( $ts, 12, 2 );
1554 $possibleTtls = array();
1555 $timeRemainingInDay = 86400 - substr( $ts, 8, 2 ) * 3600 -
1556 substr( $ts, 10, 2 ) * 60 - substr( $ts, 12, 2 );
1559 ( 7 - Language
::dateTimeObjFormat( $dateTimeObj, $ts, $zone, 'N' ) ) * 86400 +
1560 $timeRemainingInDay;
1561 } elseif ( $usedISOYear ) {
1562 // December 28th falls on the last ISO week of the year, every year.
1563 // The last ISO week of a year can be 52 or 53.
1564 $lastWeekOfISOYear = DateTime
::createFromFormat(
1566 substr( $ts, 0, 4 ) . '1228',
1567 $zone ?
: new DateTimeZone( 'UTC' )
1569 $currentISOWeek = Language
::dateTimeObjFormat( $dateTimeObj, $ts, $zone, 'W' );
1570 $weeksRemaining = $lastWeekOfISOYear - $currentISOWeek;
1571 $timeRemainingInWeek =
1572 ( 7 - Language
::dateTimeObjFormat( $dateTimeObj, $ts, $zone, 'N' ) ) * 86400
1573 +
$timeRemainingInDay;
1574 $possibleTtls[] = $weeksRemaining * 604800 +
$timeRemainingInWeek;
1579 ( Language
::dateTimeObjFormat( $dateTimeObj, $ts, $zone, 't' ) -
1580 substr( $ts, 6, 2 ) ) * 86400
1581 +
$timeRemainingInDay;
1582 } elseif ( $usedYear ) {
1584 ( Language
::dateTimeObjFormat( $dateTimeObj, $ts, $zone, 'L' ) +
364 -
1585 Language
::dateTimeObjFormat( $dateTimeObj, $ts, $zone, 'z' ) ) * 86400
1586 +
$timeRemainingInDay;
1587 } elseif ( $usedIsLeapYear ) {
1588 $year = substr( $ts, 0, 4 );
1589 $timeRemainingInYear =
1590 ( Language
::dateTimeObjFormat( $dateTimeObj, $ts, $zone, 'L' ) +
364 -
1591 Language
::dateTimeObjFormat( $dateTimeObj, $ts, $zone, 'z' ) ) * 86400
1592 +
$timeRemainingInDay;
1594 if ( $mod ||
( !( $year %
100 ) && $year %
400 ) ) {
1595 // this isn't a leap year. see when the next one starts
1596 $nextCandidate = $year - $mod +
4;
1597 if ( $nextCandidate %
100 ||
!( $nextCandidate %
400 ) ) {
1598 $possibleTtls[] = ( $nextCandidate - $year - 1 ) * 365 * 86400 +
1599 $timeRemainingInYear;
1601 $possibleTtls[] = ( $nextCandidate - $year +
3 ) * 365 * 86400 +
1602 $timeRemainingInYear;
1605 // this is a leap year, so the next year isn't
1606 $possibleTtls[] = $timeRemainingInYear;
1610 if ( $possibleTtls ) {
1611 $ttl = min( $possibleTtls );
1618 private static $GREG_DAYS = array( 31, 28, 31, 30, 31, 30, 31, 31, 30, 31, 30, 31 );
1619 private static $IRANIAN_DAYS = array( 31, 31, 31, 31, 31, 31, 30, 30, 30, 30, 30, 29 );
1622 * Algorithm by Roozbeh Pournader and Mohammad Toossi to convert
1623 * Gregorian dates to Iranian dates. Originally written in C, it
1624 * is released under the terms of GNU Lesser General Public
1625 * License. Conversion to PHP was performed by Niklas Laxström.
1627 * Link: http://www.farsiweb.info/jalali/jalali.c
1633 private static function tsToIranian( $ts ) {
1634 $gy = substr( $ts, 0, 4 ) -1600;
1635 $gm = substr( $ts, 4, 2 ) -1;
1636 $gd = substr( $ts, 6, 2 ) -1;
1638 # Days passed from the beginning (including leap years)
1640 +
floor( ( $gy +
3 ) / 4 )
1641 - floor( ( $gy +
99 ) / 100 )
1642 +
floor( ( $gy +
399 ) / 400 );
1644 // Add days of the past months of this year
1645 for ( $i = 0; $i < $gm; $i++
) {
1646 $gDayNo +
= self
::$GREG_DAYS[$i];
1650 if ( $gm > 1 && ( ( $gy %
4 === 0 && $gy %
100 !== 0 ||
( $gy %
400 == 0 ) ) ) ) {
1654 // Days passed in current month
1655 $gDayNo +
= (int)$gd;
1657 $jDayNo = $gDayNo - 79;
1659 $jNp = floor( $jDayNo / 12053 );
1662 $jy = 979 +
33 * $jNp +
4 * floor( $jDayNo / 1461 );
1665 if ( $jDayNo >= 366 ) {
1666 $jy +
= floor( ( $jDayNo - 1 ) / 365 );
1667 $jDayNo = floor( ( $jDayNo - 1 ) %
365 );
1670 for ( $i = 0; $i < 11 && $jDayNo >= self
::$IRANIAN_DAYS[$i]; $i++
) {
1671 $jDayNo -= self
::$IRANIAN_DAYS[$i];
1677 return array( $jy, $jm, $jd );
1681 * Converting Gregorian dates to Hijri dates.
1683 * Based on a PHP-Nuke block by Sharjeel which is released under GNU/GPL license
1685 * @see http://phpnuke.org/modules.php?name=News&file=article&sid=8234&mode=thread&order=0&thold=0
1691 private static function tsToHijri( $ts ) {
1692 $year = substr( $ts, 0, 4 );
1693 $month = substr( $ts, 4, 2 );
1694 $day = substr( $ts, 6, 2 );
1702 ( $zy > 1582 ) ||
( ( $zy == 1582 ) && ( $zm > 10 ) ) ||
1703 ( ( $zy == 1582 ) && ( $zm == 10 ) && ( $zd > 14 ) )
1705 $zjd = (int)( ( 1461 * ( $zy +
4800 +
(int)( ( $zm - 14 ) / 12 ) ) ) / 4 ) +
1706 (int)( ( 367 * ( $zm - 2 - 12 * ( (int)( ( $zm - 14 ) / 12 ) ) ) ) / 12 ) -
1707 (int)( ( 3 * (int)( ( ( $zy +
4900 +
(int)( ( $zm - 14 ) / 12 ) ) / 100 ) ) ) / 4 ) +
1710 $zjd = 367 * $zy - (int)( ( 7 * ( $zy +
5001 +
(int)( ( $zm - 9 ) / 7 ) ) ) / 4 ) +
1711 (int)( ( 275 * $zm ) / 9 ) +
$zd +
1729777;
1714 $zl = $zjd -1948440 +
10632;
1715 $zn = (int)( ( $zl - 1 ) / 10631 );
1716 $zl = $zl - 10631 * $zn +
354;
1717 $zj = ( (int)( ( 10985 - $zl ) / 5316 ) ) * ( (int)( ( 50 * $zl ) / 17719 ) ) +
1718 ( (int)( $zl / 5670 ) ) * ( (int)( ( 43 * $zl ) / 15238 ) );
1719 $zl = $zl - ( (int)( ( 30 - $zj ) / 15 ) ) * ( (int)( ( 17719 * $zj ) / 50 ) ) -
1720 ( (int)( $zj / 16 ) ) * ( (int)( ( 15238 * $zj ) / 43 ) ) +
29;
1721 $zm = (int)( ( 24 * $zl ) / 709 );
1722 $zd = $zl - (int)( ( 709 * $zm ) / 24 );
1723 $zy = 30 * $zn +
$zj - 30;
1725 return array( $zy, $zm, $zd );
1729 * Converting Gregorian dates to Hebrew dates.
1731 * Based on a JavaScript code by Abu Mami and Yisrael Hersch
1732 * (abu-mami@kaluach.net, http://www.kaluach.net), who permitted
1733 * to translate the relevant functions into PHP and release them under
1736 * The months are counted from Tishrei = 1. In a leap year, Adar I is 13
1737 * and Adar II is 14. In a non-leap year, Adar is 6.
1743 private static function tsToHebrew( $ts ) {
1745 $year = substr( $ts, 0, 4 );
1746 $month = substr( $ts, 4, 2 );
1747 $day = substr( $ts, 6, 2 );
1749 # Calculate Hebrew year
1750 $hebrewYear = $year +
3760;
1752 # Month number when September = 1, August = 12
1754 if ( $month > 12 ) {
1761 # Calculate day of year from 1 September
1763 for ( $i = 1; $i < $month; $i++
) {
1767 # Check if the year is leap
1768 if ( $year %
400 == 0 ||
( $year %
4 == 0 && $year %
100 > 0 ) ) {
1771 } elseif ( $i == 8 ||
$i == 10 ||
$i == 1 ||
$i == 3 ) {
1778 # Calculate the start of the Hebrew year
1779 $start = self
::hebrewYearStart( $hebrewYear );
1781 # Calculate next year's start
1782 if ( $dayOfYear <= $start ) {
1783 # Day is before the start of the year - it is the previous year
1785 $nextStart = $start;
1789 # Add days since previous year's 1 September
1791 if ( ( $year %
400 == 0 ) ||
( $year %
100 != 0 && $year %
4 == 0 ) ) {
1795 # Start of the new (previous) year
1796 $start = self
::hebrewYearStart( $hebrewYear );
1799 $nextStart = self
::hebrewYearStart( $hebrewYear +
1 );
1802 # Calculate Hebrew day of year
1803 $hebrewDayOfYear = $dayOfYear - $start;
1805 # Difference between year's days
1806 $diff = $nextStart - $start;
1807 # Add 12 (or 13 for leap years) days to ignore the difference between
1808 # Hebrew and Gregorian year (353 at least vs. 365/6) - now the
1809 # difference is only about the year type
1810 if ( ( $year %
400 == 0 ) ||
( $year %
100 != 0 && $year %
4 == 0 ) ) {
1816 # Check the year pattern, and is leap year
1817 # 0 means an incomplete year, 1 means a regular year, 2 means a complete year
1818 # This is mod 30, to work on both leap years (which add 30 days of Adar I)
1819 # and non-leap years
1820 $yearPattern = $diff %
30;
1821 # Check if leap year
1822 $isLeap = $diff >= 30;
1824 # Calculate day in the month from number of day in the Hebrew year
1825 # Don't check Adar - if the day is not in Adar, we will stop before;
1826 # if it is in Adar, we will use it to check if it is Adar I or Adar II
1827 $hebrewDay = $hebrewDayOfYear;
1830 while ( $hebrewMonth <= 12 ) {
1831 # Calculate days in this month
1832 if ( $isLeap && $hebrewMonth == 6 ) {
1833 # Adar in a leap year
1835 # Leap year - has Adar I, with 30 days, and Adar II, with 29 days
1837 if ( $hebrewDay <= $days ) {
1841 # Subtract the days of Adar I
1842 $hebrewDay -= $days;
1845 if ( $hebrewDay <= $days ) {
1851 } elseif ( $hebrewMonth == 2 && $yearPattern == 2 ) {
1852 # Cheshvan in a complete year (otherwise as the rule below)
1854 } elseif ( $hebrewMonth == 3 && $yearPattern == 0 ) {
1855 # Kislev in an incomplete year (otherwise as the rule below)
1858 # Odd months have 30 days, even have 29
1859 $days = 30 - ( $hebrewMonth - 1 ) %
2;
1861 if ( $hebrewDay <= $days ) {
1862 # In the current month
1865 # Subtract the days of the current month
1866 $hebrewDay -= $days;
1867 # Try in the next month
1872 return array( $hebrewYear, $hebrewMonth, $hebrewDay, $days );
1876 * This calculates the Hebrew year start, as days since 1 September.
1877 * Based on Carl Friedrich Gauss algorithm for finding Easter date.
1878 * Used for Hebrew date.
1884 private static function hebrewYearStart( $year ) {
1885 $a = intval( ( 12 * ( $year - 1 ) +
17 ) %
19 );
1886 $b = intval( ( $year - 1 ) %
4 );
1887 $m = 32.044093161144 +
1.5542417966212 * $a +
$b / 4.0 - 0.0031777940220923 * ( $year - 1 );
1891 $Mar = intval( $m );
1897 $c = intval( ( $Mar +
3 * ( $year - 1 ) +
5 * $b +
5 ) %
7 );
1898 if ( $c == 0 && $a > 11 && $m >= 0.89772376543210 ) {
1900 } elseif ( $c == 1 && $a > 6 && $m >= 0.63287037037037 ) {
1902 } elseif ( $c == 2 ||
$c == 4 ||
$c == 6 ) {
1906 $Mar +
= intval( ( $year - 3761 ) / 100 ) - intval( ( $year - 3761 ) / 400 ) - 24;
1911 * Algorithm to convert Gregorian dates to Thai solar dates,
1912 * Minguo dates or Minguo dates.
1914 * Link: http://en.wikipedia.org/wiki/Thai_solar_calendar
1915 * http://en.wikipedia.org/wiki/Minguo_calendar
1916 * http://en.wikipedia.org/wiki/Japanese_era_name
1918 * @param string $ts 14-character timestamp
1919 * @param string $cName Calender name
1920 * @return array Converted year, month, day
1922 private static function tsToYear( $ts, $cName ) {
1923 $gy = substr( $ts, 0, 4 );
1924 $gm = substr( $ts, 4, 2 );
1925 $gd = substr( $ts, 6, 2 );
1927 if ( !strcmp( $cName, 'thai' ) ) {
1929 # Add 543 years to the Gregorian calendar
1930 # Months and days are identical
1931 $gy_offset = $gy +
543;
1932 } elseif ( ( !strcmp( $cName, 'minguo' ) ) ||
!strcmp( $cName, 'juche' ) ) {
1934 # Deduct 1911 years from the Gregorian calendar
1935 # Months and days are identical
1936 $gy_offset = $gy - 1911;
1937 } elseif ( !strcmp( $cName, 'tenno' ) ) {
1938 # Nengō dates up to Meiji period
1939 # Deduct years from the Gregorian calendar
1940 # depending on the nengo periods
1941 # Months and days are identical
1943 ||
( ( $gy == 1912 ) && ( $gm < 7 ) )
1944 ||
( ( $gy == 1912 ) && ( $gm == 7 ) && ( $gd < 31 ) )
1947 $gy_gannen = $gy - 1868 +
1;
1948 $gy_offset = $gy_gannen;
1949 if ( $gy_gannen == 1 ) {
1952 $gy_offset = '明治' . $gy_offset;
1954 ( ( $gy == 1912 ) && ( $gm == 7 ) && ( $gd == 31 ) ) ||
1955 ( ( $gy == 1912 ) && ( $gm >= 8 ) ) ||
1956 ( ( $gy > 1912 ) && ( $gy < 1926 ) ) ||
1957 ( ( $gy == 1926 ) && ( $gm < 12 ) ) ||
1958 ( ( $gy == 1926 ) && ( $gm == 12 ) && ( $gd < 26 ) )
1961 $gy_gannen = $gy - 1912 +
1;
1962 $gy_offset = $gy_gannen;
1963 if ( $gy_gannen == 1 ) {
1966 $gy_offset = '大正' . $gy_offset;
1968 ( ( $gy == 1926 ) && ( $gm == 12 ) && ( $gd >= 26 ) ) ||
1969 ( ( $gy > 1926 ) && ( $gy < 1989 ) ) ||
1970 ( ( $gy == 1989 ) && ( $gm == 1 ) && ( $gd < 8 ) )
1973 $gy_gannen = $gy - 1926 +
1;
1974 $gy_offset = $gy_gannen;
1975 if ( $gy_gannen == 1 ) {
1978 $gy_offset = '昭和' . $gy_offset;
1981 $gy_gannen = $gy - 1989 +
1;
1982 $gy_offset = $gy_gannen;
1983 if ( $gy_gannen == 1 ) {
1986 $gy_offset = '平成' . $gy_offset;
1992 return array( $gy_offset, $gm, $gd );
1996 * Roman number formatting up to 10000
2002 static function romanNumeral( $num ) {
2003 static $table = array(
2004 array( '', 'I', 'II', 'III', 'IV', 'V', 'VI', 'VII', 'VIII', 'IX', 'X' ),
2005 array( '', 'X', 'XX', 'XXX', 'XL', 'L', 'LX', 'LXX', 'LXXX', 'XC', 'C' ),
2006 array( '', 'C', 'CC', 'CCC', 'CD', 'D', 'DC', 'DCC', 'DCCC', 'CM', 'M' ),
2007 array( '', 'M', 'MM', 'MMM', 'MMMM', 'MMMMM', 'MMMMMM', 'MMMMMMM',
2008 'MMMMMMMM', 'MMMMMMMMM', 'MMMMMMMMMM' )
2011 $num = intval( $num );
2012 if ( $num > 10000 ||
$num <= 0 ) {
2017 for ( $pow10 = 1000, $i = 3; $i >= 0; $pow10 /= 10, $i-- ) {
2018 if ( $num >= $pow10 ) {
2019 $s .= $table[$i][(int)floor( $num / $pow10 )];
2021 $num = $num %
$pow10;
2027 * Hebrew Gematria number formatting up to 9999
2033 static function hebrewNumeral( $num ) {
2034 static $table = array(
2035 array( '', 'א', 'ב', 'ג', 'ד', 'ה', 'ו', 'ז', 'ח', 'ט', 'י' ),
2036 array( '', 'י', 'כ', 'ל', 'מ', 'נ', 'ס', 'ע', 'פ', 'צ', 'ק' ),
2046 array( 'ת', 'ת', 'ק' ),
2047 array( 'ת', 'ת', 'ר' ),
2049 array( '', 'א', 'ב', 'ג', 'ד', 'ה', 'ו', 'ז', 'ח', 'ט', 'י' )
2052 $num = intval( $num );
2053 if ( $num > 9999 ||
$num <= 0 ) {
2057 // Round thousands have special notations
2058 if ( $num === 1000 ) {
2060 } elseif ( $num %
1000 === 0 ) {
2061 return $table[0][$num / 1000] . "' אלפים";
2066 for ( $pow10 = 1000, $i = 3; $i >= 0; $pow10 /= 10, $i-- ) {
2067 if ( $num >= $pow10 ) {
2068 if ( $num === 15 ||
$num === 16 ) {
2069 $letters[] = $table[0][9];
2070 $letters[] = $table[0][$num - 9];
2073 $letters = array_merge(
2075 (array)$table[$i][intval( $num / $pow10 )]
2078 if ( $pow10 === 1000 ) {
2084 $num = $num %
$pow10;
2087 $preTransformLength = count( $letters );
2088 if ( $preTransformLength === 1 ) {
2089 // Add geresh (single quote) to one-letter numbers
2092 $lastIndex = $preTransformLength - 1;
2093 $letters[$lastIndex] = str_replace(
2094 array( 'כ', 'מ', 'נ', 'פ', 'צ' ),
2095 array( 'ך', 'ם', 'ן', 'ף', 'ץ' ),
2096 $letters[$lastIndex]
2099 // Add gershayim (double quote) to multiple-letter numbers,
2100 // but exclude numbers with only one letter after the thousands
2101 // (1001-1009, 1020, 1030, 2001-2009, etc.)
2102 if ( $letters[1] === "'" && $preTransformLength === 3 ) {
2105 array_splice( $letters, -1, 0, '"' );
2109 return implode( $letters );
2113 * Used by date() and time() to adjust the time output.
2115 * @param string $ts The time in date('YmdHis') format
2116 * @param mixed $tz Adjust the time by this amount (default false, mean we
2117 * get user timecorrection setting)
2120 function userAdjust( $ts, $tz = false ) {
2121 global $wgUser, $wgLocalTZoffset;
2123 if ( $tz === false ) {
2124 $tz = $wgUser->getOption( 'timecorrection' );
2127 $data = explode( '|', $tz, 3 );
2129 if ( $data[0] == 'ZoneInfo' ) {
2130 wfSuppressWarnings();
2131 $userTZ = timezone_open( $data[2] );
2132 wfRestoreWarnings();
2133 if ( $userTZ !== false ) {
2134 $date = date_create( $ts, timezone_open( 'UTC' ) );
2135 date_timezone_set( $date, $userTZ );
2136 $date = date_format( $date, 'YmdHis' );
2139 # Unrecognized timezone, default to 'Offset' with the stored offset.
2140 $data[0] = 'Offset';
2143 if ( $data[0] == 'System' ||
$tz == '' ) {
2144 # Global offset in minutes.
2145 $minDiff = $wgLocalTZoffset;
2146 } elseif ( $data[0] == 'Offset' ) {
2147 $minDiff = intval( $data[1] );
2149 $data = explode( ':', $tz );
2150 if ( count( $data ) == 2 ) {
2151 $data[0] = intval( $data[0] );
2152 $data[1] = intval( $data[1] );
2153 $minDiff = abs( $data[0] ) * 60 +
$data[1];
2154 if ( $data[0] < 0 ) {
2155 $minDiff = -$minDiff;
2158 $minDiff = intval( $data[0] ) * 60;
2162 # No difference ? Return time unchanged
2163 if ( 0 == $minDiff ) {
2167 wfSuppressWarnings(); // E_STRICT system time bitching
2168 # Generate an adjusted date; take advantage of the fact that mktime
2169 # will normalize out-of-range values so we don't have to split $minDiff
2170 # into hours and minutes.
2172 (int)substr( $ts, 8, 2 ) ), # Hours
2173 (int)substr( $ts, 10, 2 ) +
$minDiff, # Minutes
2174 (int)substr( $ts, 12, 2 ), # Seconds
2175 (int)substr( $ts, 4, 2 ), # Month
2176 (int)substr( $ts, 6, 2 ), # Day
2177 (int)substr( $ts, 0, 4 ) ); # Year
2179 $date = date( 'YmdHis', $t );
2180 wfRestoreWarnings();
2186 * This is meant to be used by time(), date(), and timeanddate() to get
2187 * the date preference they're supposed to use, it should be used in
2191 * function timeanddate([...], $format = true) {
2192 * $datePreference = $this->dateFormat($format);
2197 * @param int|string|bool $usePrefs If true, the user's preference is used
2198 * if false, the site/language default is used
2199 * if int/string, assumed to be a format.
2202 function dateFormat( $usePrefs = true ) {
2205 if ( is_bool( $usePrefs ) ) {
2207 $datePreference = $wgUser->getDatePreference();
2209 $datePreference = (string)User
::getDefaultOption( 'date' );
2212 $datePreference = (string)$usePrefs;
2216 if ( $datePreference == '' ) {
2220 return $datePreference;
2224 * Get a format string for a given type and preference
2225 * @param string $type May be date, time or both
2226 * @param string $pref The format name as it appears in Messages*.php
2228 * @since 1.22 New type 'pretty' that provides a more readable timestamp format
2232 function getDateFormatString( $type, $pref ) {
2233 if ( !isset( $this->dateFormatStrings
[$type][$pref] ) ) {
2234 if ( $pref == 'default' ) {
2235 $pref = $this->getDefaultDateFormat();
2236 $df = self
::$dataCache->getSubitem( $this->mCode
, 'dateFormats', "$pref $type" );
2238 $df = self
::$dataCache->getSubitem( $this->mCode
, 'dateFormats', "$pref $type" );
2240 if ( $type === 'pretty' && $df === null ) {
2241 $df = $this->getDateFormatString( 'date', $pref );
2244 if ( $df === null ) {
2245 $pref = $this->getDefaultDateFormat();
2246 $df = self
::$dataCache->getSubitem( $this->mCode
, 'dateFormats', "$pref $type" );
2249 $this->dateFormatStrings
[$type][$pref] = $df;
2251 return $this->dateFormatStrings
[$type][$pref];
2255 * @param string $ts The time format which needs to be turned into a
2256 * date('YmdHis') format with wfTimestamp(TS_MW,$ts)
2257 * @param bool $adj Whether to adjust the time output according to the
2258 * user configured offset ($timecorrection)
2259 * @param mixed $format True to use user's date format preference
2260 * @param string|bool $timecorrection The time offset as returned by
2261 * validateTimeZone() in Special:Preferences
2264 function date( $ts, $adj = false, $format = true, $timecorrection = false ) {
2265 $ts = wfTimestamp( TS_MW
, $ts );
2267 $ts = $this->userAdjust( $ts, $timecorrection );
2269 $df = $this->getDateFormatString( 'date', $this->dateFormat( $format ) );
2270 return $this->sprintfDate( $df, $ts );
2274 * @param string $ts The time format which needs to be turned into a
2275 * date('YmdHis') format with wfTimestamp(TS_MW,$ts)
2276 * @param bool $adj Whether to adjust the time output according to the
2277 * user configured offset ($timecorrection)
2278 * @param mixed $format True to use user's date format preference
2279 * @param string|bool $timecorrection The time offset as returned by
2280 * validateTimeZone() in Special:Preferences
2283 function time( $ts, $adj = false, $format = true, $timecorrection = false ) {
2284 $ts = wfTimestamp( TS_MW
, $ts );
2286 $ts = $this->userAdjust( $ts, $timecorrection );
2288 $df = $this->getDateFormatString( 'time', $this->dateFormat( $format ) );
2289 return $this->sprintfDate( $df, $ts );
2293 * @param string $ts The time format which needs to be turned into a
2294 * date('YmdHis') format with wfTimestamp(TS_MW,$ts)
2295 * @param bool $adj Whether to adjust the time output according to the
2296 * user configured offset ($timecorrection)
2297 * @param mixed $format What format to return, if it's false output the
2298 * default one (default true)
2299 * @param string|bool $timecorrection The time offset as returned by
2300 * validateTimeZone() in Special:Preferences
2303 function timeanddate( $ts, $adj = false, $format = true, $timecorrection = false ) {
2304 $ts = wfTimestamp( TS_MW
, $ts );
2306 $ts = $this->userAdjust( $ts, $timecorrection );
2308 $df = $this->getDateFormatString( 'both', $this->dateFormat( $format ) );
2309 return $this->sprintfDate( $df, $ts );
2313 * Takes a number of seconds and turns it into a text using values such as hours and minutes.
2317 * @param int $seconds The amount of seconds.
2318 * @param array $chosenIntervals The intervals to enable.
2322 public function formatDuration( $seconds, array $chosenIntervals = array() ) {
2323 $intervals = $this->getDurationIntervals( $seconds, $chosenIntervals );
2325 $segments = array();
2327 foreach ( $intervals as $intervalName => $intervalValue ) {
2328 // Messages: duration-seconds, duration-minutes, duration-hours, duration-days, duration-weeks,
2329 // duration-years, duration-decades, duration-centuries, duration-millennia
2330 $message = wfMessage( 'duration-' . $intervalName )->numParams( $intervalValue );
2331 $segments[] = $message->inLanguage( $this )->escaped();
2334 return $this->listToText( $segments );
2338 * Takes a number of seconds and returns an array with a set of corresponding intervals.
2339 * For example 65 will be turned into array( minutes => 1, seconds => 5 ).
2343 * @param int $seconds The amount of seconds.
2344 * @param array $chosenIntervals The intervals to enable.
2348 public function getDurationIntervals( $seconds, array $chosenIntervals = array() ) {
2349 if ( empty( $chosenIntervals ) ) {
2350 $chosenIntervals = array(
2362 $intervals = array_intersect_key( self
::$durationIntervals, array_flip( $chosenIntervals ) );
2363 $sortedNames = array_keys( $intervals );
2364 $smallestInterval = array_pop( $sortedNames );
2366 $segments = array();
2368 foreach ( $intervals as $name => $length ) {
2369 $value = floor( $seconds / $length );
2371 if ( $value > 0 ||
( $name == $smallestInterval && empty( $segments ) ) ) {
2372 $seconds -= $value * $length;
2373 $segments[$name] = $value;
2381 * Internal helper function for userDate(), userTime() and userTimeAndDate()
2383 * @param string $type Can be 'date', 'time' or 'both'
2384 * @param string $ts The time format which needs to be turned into a
2385 * date('YmdHis') format with wfTimestamp(TS_MW,$ts)
2386 * @param User $user User object used to get preferences for timezone and format
2387 * @param array $options Array, can contain the following keys:
2388 * - 'timecorrection': time correction, can have the following values:
2389 * - true: use user's preference
2390 * - false: don't use time correction
2391 * - int: value of time correction in minutes
2392 * - 'format': format to use, can have the following values:
2393 * - true: use user's preference
2394 * - false: use default preference
2395 * - string: format to use
2399 private function internalUserTimeAndDate( $type, $ts, User
$user, array $options ) {
2400 $ts = wfTimestamp( TS_MW
, $ts );
2401 $options +
= array( 'timecorrection' => true, 'format' => true );
2402 if ( $options['timecorrection'] !== false ) {
2403 if ( $options['timecorrection'] === true ) {
2404 $offset = $user->getOption( 'timecorrection' );
2406 $offset = $options['timecorrection'];
2408 $ts = $this->userAdjust( $ts, $offset );
2410 if ( $options['format'] === true ) {
2411 $format = $user->getDatePreference();
2413 $format = $options['format'];
2415 $df = $this->getDateFormatString( $type, $this->dateFormat( $format ) );
2416 return $this->sprintfDate( $df, $ts );
2420 * Get the formatted date for the given timestamp and formatted for
2423 * @param mixed $ts Mixed: the time format which needs to be turned into a
2424 * date('YmdHis') format with wfTimestamp(TS_MW,$ts)
2425 * @param User $user User object used to get preferences for timezone and format
2426 * @param array $options Array, can contain the following keys:
2427 * - 'timecorrection': time correction, can have the following values:
2428 * - true: use user's preference
2429 * - false: don't use time correction
2430 * - int: value of time correction in minutes
2431 * - 'format': format to use, can have the following values:
2432 * - true: use user's preference
2433 * - false: use default preference
2434 * - string: format to use
2438 public function userDate( $ts, User
$user, array $options = array() ) {
2439 return $this->internalUserTimeAndDate( 'date', $ts, $user, $options );
2443 * Get the formatted time for the given timestamp and formatted for
2446 * @param mixed $ts The time format which needs to be turned into a
2447 * date('YmdHis') format with wfTimestamp(TS_MW,$ts)
2448 * @param User $user User object used to get preferences for timezone and format
2449 * @param array $options Array, can contain the following keys:
2450 * - 'timecorrection': time correction, can have the following values:
2451 * - true: use user's preference
2452 * - false: don't use time correction
2453 * - int: value of time correction in minutes
2454 * - 'format': format to use, can have the following values:
2455 * - true: use user's preference
2456 * - false: use default preference
2457 * - string: format to use
2461 public function userTime( $ts, User
$user, array $options = array() ) {
2462 return $this->internalUserTimeAndDate( 'time', $ts, $user, $options );
2466 * Get the formatted date and time for the given timestamp and formatted for
2469 * @param mixed $ts The time format which needs to be turned into a
2470 * date('YmdHis') format with wfTimestamp(TS_MW,$ts)
2471 * @param User $user User object used to get preferences for timezone and format
2472 * @param array $options Array, can contain the following keys:
2473 * - 'timecorrection': time correction, can have the following values:
2474 * - true: use user's preference
2475 * - false: don't use time correction
2476 * - int: value of time correction in minutes
2477 * - 'format': format to use, can have the following values:
2478 * - true: use user's preference
2479 * - false: use default preference
2480 * - string: format to use
2484 public function userTimeAndDate( $ts, User
$user, array $options = array() ) {
2485 return $this->internalUserTimeAndDate( 'both', $ts, $user, $options );
2489 * Get the timestamp in a human-friendly relative format, e.g., "3 days ago".
2491 * Determine the difference between the timestamp and the current time, and
2492 * generate a readable timestamp by returning "<N> <units> ago", where the
2493 * largest possible unit is used.
2495 * @since 1.26 (Prior to 1.26 method existed but was not meant to be used directly)
2497 * @param MWTimestamp $time
2498 * @param MWTimestamp|null $relativeTo The base timestamp to compare to (defaults to now)
2499 * @param User|null $user User the timestamp is being generated for (or null to use main context's user)
2500 * @return string Formatted timestamp
2502 public function getHumanTimestamp( MWTimestamp
$time, MWTimestamp
$relativeTo = null, User
$user = null ) {
2503 if ( $relativeTo === null ) {
2504 $relativeTo = new MWTimestamp();
2506 if ( $user === null ) {
2507 $user = RequestContext
::getMain()->getUser();
2510 // Adjust for the user's timezone.
2511 $offsetThis = $time->offsetForUser( $user );
2512 $offsetRel = $relativeTo->offsetForUser( $user );
2515 if ( Hooks
::run( 'GetHumanTimestamp', array( &$ts, $time, $relativeTo, $user, $this ) ) ) {
2516 $ts = $this->getHumanTimestampInternal( $time, $relativeTo, $user );
2519 // Reset the timezone on the objects.
2520 $time->timestamp
->sub( $offsetThis );
2521 $relativeTo->timestamp
->sub( $offsetRel );
2527 * Convert an MWTimestamp into a pretty human-readable timestamp using
2528 * the given user preferences and relative base time.
2530 * @see Language::getHumanTimestamp
2531 * @param MWTimestamp $ts Timestamp to prettify
2532 * @param MWTimestamp $relativeTo Base timestamp
2533 * @param User $user User preferences to use
2534 * @return string Human timestamp
2537 private function getHumanTimestampInternal( MWTimestamp
$ts, MWTimestamp
$relativeTo, User
$user ) {
2538 $diff = $ts->diff( $relativeTo );
2539 $diffDay = (bool)( (int)$ts->timestamp
->format( 'w' ) -
2540 (int)$relativeTo->timestamp
->format( 'w' ) );
2541 $days = $diff->days ?
: (int)$diffDay;
2542 if ( $diff->invert ||
$days > 5
2543 && $ts->timestamp
->format( 'Y' ) !== $relativeTo->timestamp
->format( 'Y' )
2545 // Timestamps are in different years: use full timestamp
2546 // Also do full timestamp for future dates
2548 * @todo FIXME: Add better handling of future timestamps.
2550 $format = $this->getDateFormatString( 'both', $user->getDatePreference() ?
: 'default' );
2551 $ts = $this->sprintfDate( $format, $ts->getTimestamp( TS_MW
) );
2552 } elseif ( $days > 5 ) {
2553 // Timestamps are in same year, but more than 5 days ago: show day and month only.
2554 $format = $this->getDateFormatString( 'pretty', $user->getDatePreference() ?
: 'default' );
2555 $ts = $this->sprintfDate( $format, $ts->getTimestamp( TS_MW
) );
2556 } elseif ( $days > 1 ) {
2557 // Timestamp within the past week: show the day of the week and time
2558 $format = $this->getDateFormatString( 'time', $user->getDatePreference() ?
: 'default' );
2559 $weekday = self
::$mWeekdayMsgs[$ts->timestamp
->format( 'w' )];
2561 // sunday-at, monday-at, tuesday-at, wednesday-at, thursday-at, friday-at, saturday-at
2562 $ts = wfMessage( "$weekday-at" )
2563 ->inLanguage( $this )
2564 ->params( $this->sprintfDate( $format, $ts->getTimestamp( TS_MW
) ) )
2566 } elseif ( $days == 1 ) {
2567 // Timestamp was yesterday: say 'yesterday' and the time.
2568 $format = $this->getDateFormatString( 'time', $user->getDatePreference() ?
: 'default' );
2569 $ts = wfMessage( 'yesterday-at' )
2570 ->inLanguage( $this )
2571 ->params( $this->sprintfDate( $format, $ts->getTimestamp( TS_MW
) ) )
2573 } elseif ( $diff->h
> 1 ||
$diff->h
== 1 && $diff->i
> 30 ) {
2574 // Timestamp was today, but more than 90 minutes ago: say 'today' and the time.
2575 $format = $this->getDateFormatString( 'time', $user->getDatePreference() ?
: 'default' );
2576 $ts = wfMessage( 'today-at' )
2577 ->inLanguage( $this )
2578 ->params( $this->sprintfDate( $format, $ts->getTimestamp( TS_MW
) ) )
2581 // From here on in, the timestamp was soon enough ago so that we can simply say
2582 // XX units ago, e.g., "2 hours ago" or "5 minutes ago"
2583 } elseif ( $diff->h
== 1 ) {
2584 // Less than 90 minutes, but more than an hour ago.
2585 $ts = wfMessage( 'hours-ago' )->inLanguage( $this )->numParams( 1 )->text();
2586 } elseif ( $diff->i
>= 1 ) {
2587 // A few minutes ago.
2588 $ts = wfMessage( 'minutes-ago' )->inLanguage( $this )->numParams( $diff->i
)->text();
2589 } elseif ( $diff->s
>= 30 ) {
2590 // Less than a minute, but more than 30 sec ago.
2591 $ts = wfMessage( 'seconds-ago' )->inLanguage( $this )->numParams( $diff->s
)->text();
2593 // Less than 30 seconds ago.
2594 $ts = wfMessage( 'just-now' )->text();
2601 * @param string $key
2602 * @return array|null
2604 function getMessage( $key ) {
2605 return self
::$dataCache->getSubitem( $this->mCode
, 'messages', $key );
2611 function getAllMessages() {
2612 return self
::$dataCache->getItem( $this->mCode
, 'messages' );
2617 * @param string $out
2618 * @param string $string
2621 function iconv( $in, $out, $string ) {
2622 # This is a wrapper for iconv in all languages except esperanto,
2623 # which does some nasty x-conversions beforehand
2625 # Even with //IGNORE iconv can whine about illegal characters in
2626 # *input* string. We just ignore those too.
2627 # REF: http://bugs.php.net/bug.php?id=37166
2628 # REF: https://bugzilla.wikimedia.org/show_bug.cgi?id=16885
2629 wfSuppressWarnings();
2630 $text = iconv( $in, $out . '//IGNORE', $string );
2631 wfRestoreWarnings();
2635 // callback functions for uc(), lc(), ucwords(), ucwordbreaks()
2638 * @param array $matches
2639 * @return mixed|string
2641 function ucwordbreaksCallbackAscii( $matches ) {
2642 return $this->ucfirst( $matches[1] );
2646 * @param array $matches
2649 function ucwordbreaksCallbackMB( $matches ) {
2650 return mb_strtoupper( $matches[0] );
2654 * @param array $matches
2657 function ucCallback( $matches ) {
2658 list( $wikiUpperChars ) = self
::getCaseMaps();
2659 return strtr( $matches[1], $wikiUpperChars );
2663 * @param array $matches
2666 function lcCallback( $matches ) {
2667 list( , $wikiLowerChars ) = self
::getCaseMaps();
2668 return strtr( $matches[1], $wikiLowerChars );
2672 * @param array $matches
2675 function ucwordsCallbackMB( $matches ) {
2676 return mb_strtoupper( $matches[0] );
2680 * @param array $matches
2683 function ucwordsCallbackWiki( $matches ) {
2684 list( $wikiUpperChars ) = self
::getCaseMaps();
2685 return strtr( $matches[0], $wikiUpperChars );
2689 * Make a string's first character uppercase
2691 * @param string $str
2695 function ucfirst( $str ) {
2697 if ( $o < 96 ) { // if already uppercase...
2699 } elseif ( $o < 128 ) {
2700 return ucfirst( $str ); // use PHP's ucfirst()
2702 // fall back to more complex logic in case of multibyte strings
2703 return $this->uc( $str, true );
2708 * Convert a string to uppercase
2710 * @param string $str
2711 * @param bool $first
2715 function uc( $str, $first = false ) {
2716 if ( function_exists( 'mb_strtoupper' ) ) {
2718 if ( $this->isMultibyte( $str ) ) {
2719 return mb_strtoupper( mb_substr( $str, 0, 1 ) ) . mb_substr( $str, 1 );
2721 return ucfirst( $str );
2724 return $this->isMultibyte( $str ) ?
mb_strtoupper( $str ) : strtoupper( $str );
2727 if ( $this->isMultibyte( $str ) ) {
2728 $x = $first ?
'^' : '';
2729 return preg_replace_callback(
2730 "/$x([a-z]|[\\xc0-\\xff][\\x80-\\xbf]*)/",
2731 array( $this, 'ucCallback' ),
2735 return $first ?
ucfirst( $str ) : strtoupper( $str );
2741 * @param string $str
2742 * @return mixed|string
2744 function lcfirst( $str ) {
2747 return strval( $str );
2748 } elseif ( $o >= 128 ) {
2749 return $this->lc( $str, true );
2750 } elseif ( $o > 96 ) {
2753 $str[0] = strtolower( $str[0] );
2759 * @param string $str
2760 * @param bool $first
2761 * @return mixed|string
2763 function lc( $str, $first = false ) {
2764 if ( function_exists( 'mb_strtolower' ) ) {
2766 if ( $this->isMultibyte( $str ) ) {
2767 return mb_strtolower( mb_substr( $str, 0, 1 ) ) . mb_substr( $str, 1 );
2769 return strtolower( substr( $str, 0, 1 ) ) . substr( $str, 1 );
2772 return $this->isMultibyte( $str ) ?
mb_strtolower( $str ) : strtolower( $str );
2775 if ( $this->isMultibyte( $str ) ) {
2776 $x = $first ?
'^' : '';
2777 return preg_replace_callback(
2778 "/$x([A-Z]|[\\xc0-\\xff][\\x80-\\xbf]*)/",
2779 array( $this, 'lcCallback' ),
2783 return $first ?
strtolower( substr( $str, 0, 1 ) ) . substr( $str, 1 ) : strtolower( $str );
2789 * @param string $str
2792 function isMultibyte( $str ) {
2793 return (bool)preg_match( '/[\x80-\xff]/', $str );
2797 * @param string $str
2798 * @return mixed|string
2800 function ucwords( $str ) {
2801 if ( $this->isMultibyte( $str ) ) {
2802 $str = $this->lc( $str );
2804 // regexp to find first letter in each word (i.e. after each space)
2805 $replaceRegexp = "/^([a-z]|[\\xc0-\\xff][\\x80-\\xbf]*)| ([a-z]|[\\xc0-\\xff][\\x80-\\xbf]*)/";
2807 // function to use to capitalize a single char
2808 if ( function_exists( 'mb_strtoupper' ) ) {
2809 return preg_replace_callback(
2811 array( $this, 'ucwordsCallbackMB' ),
2815 return preg_replace_callback(
2817 array( $this, 'ucwordsCallbackWiki' ),
2822 return ucwords( strtolower( $str ) );
2827 * capitalize words at word breaks
2829 * @param string $str
2832 function ucwordbreaks( $str ) {
2833 if ( $this->isMultibyte( $str ) ) {
2834 $str = $this->lc( $str );
2836 // since \b doesn't work for UTF-8, we explicitely define word break chars
2837 $breaks = "[ \-\(\)\}\{\.,\?!]";
2839 // find first letter after word break
2840 $replaceRegexp = "/^([a-z]|[\\xc0-\\xff][\\x80-\\xbf]*)|" .
2841 "$breaks([a-z]|[\\xc0-\\xff][\\x80-\\xbf]*)/";
2843 if ( function_exists( 'mb_strtoupper' ) ) {
2844 return preg_replace_callback(
2846 array( $this, 'ucwordbreaksCallbackMB' ),
2850 return preg_replace_callback(
2852 array( $this, 'ucwordsCallbackWiki' ),
2857 return preg_replace_callback(
2858 '/\b([\w\x80-\xff]+)\b/',
2859 array( $this, 'ucwordbreaksCallbackAscii' ),
2866 * Return a case-folded representation of $s
2868 * This is a representation such that caseFold($s1)==caseFold($s2) if $s1
2869 * and $s2 are the same except for the case of their characters. It is not
2870 * necessary for the value returned to make sense when displayed.
2872 * Do *not* perform any other normalisation in this function. If a caller
2873 * uses this function when it should be using a more general normalisation
2874 * function, then fix the caller.
2880 function caseFold( $s ) {
2881 return $this->uc( $s );
2888 function checkTitleEncoding( $s ) {
2889 if ( is_array( $s ) ) {
2890 throw new MWException( 'Given array to checkTitleEncoding.' );
2892 if ( StringUtils
::isUtf8( $s ) ) {
2896 return $this->iconv( $this->fallback8bitEncoding(), 'utf-8', $s );
2902 function fallback8bitEncoding() {
2903 return self
::$dataCache->getItem( $this->mCode
, 'fallback8bitEncoding' );
2907 * Most writing systems use whitespace to break up words.
2908 * Some languages such as Chinese don't conventionally do this,
2909 * which requires special handling when breaking up words for
2914 function hasWordBreaks() {
2919 * Some languages such as Chinese require word segmentation,
2920 * Specify such segmentation when overridden in derived class.
2922 * @param string $string
2925 function segmentByWord( $string ) {
2930 * Some languages have special punctuation need to be normalized.
2931 * Make such changes here.
2933 * @param string $string
2936 function normalizeForSearch( $string ) {
2937 return self
::convertDoubleWidth( $string );
2941 * convert double-width roman characters to single-width.
2942 * range: ff00-ff5f ~= 0020-007f
2944 * @param string $string
2948 protected static function convertDoubleWidth( $string ) {
2949 static $full = null;
2950 static $half = null;
2952 if ( $full === null ) {
2953 $fullWidth = "0123456789ABCDEFGHIJKLMNOPQRSTUVWXYZabcdefghijklmnopqrstuvwxyz";
2954 $halfWidth = "0123456789ABCDEFGHIJKLMNOPQRSTUVWXYZabcdefghijklmnopqrstuvwxyz";
2955 $full = str_split( $fullWidth, 3 );
2956 $half = str_split( $halfWidth );
2959 $string = str_replace( $full, $half, $string );
2964 * @param string $string
2965 * @param string $pattern
2968 protected static function insertSpace( $string, $pattern ) {
2969 $string = preg_replace( $pattern, " $1 ", $string );
2970 $string = preg_replace( '/ +/', ' ', $string );
2975 * @param array $termsArray
2978 function convertForSearchResult( $termsArray ) {
2979 # some languages, e.g. Chinese, need to do a conversion
2980 # in order for search results to be displayed correctly
2985 * Get the first character of a string.
2990 function firstChar( $s ) {
2993 '/^([\x00-\x7f]|[\xc0-\xdf][\x80-\xbf]|' .
2994 '[\xe0-\xef][\x80-\xbf]{2}|[\xf0-\xf7][\x80-\xbf]{3})/',
2999 if ( isset( $matches[1] ) ) {
3000 if ( strlen( $matches[1] ) != 3 ) {
3004 // Break down Hangul syllables to grab the first jamo
3005 $code = UtfNormal\Utils
::utf8ToCodepoint( $matches[1] );
3006 if ( $code < 0xac00 ||
0xd7a4 <= $code ) {
3008 } elseif ( $code < 0xb098 ) {
3009 return "\xe3\x84\xb1";
3010 } elseif ( $code < 0xb2e4 ) {
3011 return "\xe3\x84\xb4";
3012 } elseif ( $code < 0xb77c ) {
3013 return "\xe3\x84\xb7";
3014 } elseif ( $code < 0xb9c8 ) {
3015 return "\xe3\x84\xb9";
3016 } elseif ( $code < 0xbc14 ) {
3017 return "\xe3\x85\x81";
3018 } elseif ( $code < 0xc0ac ) {
3019 return "\xe3\x85\x82";
3020 } elseif ( $code < 0xc544 ) {
3021 return "\xe3\x85\x85";
3022 } elseif ( $code < 0xc790 ) {
3023 return "\xe3\x85\x87";
3024 } elseif ( $code < 0xcc28 ) {
3025 return "\xe3\x85\x88";
3026 } elseif ( $code < 0xce74 ) {
3027 return "\xe3\x85\x8a";
3028 } elseif ( $code < 0xd0c0 ) {
3029 return "\xe3\x85\x8b";
3030 } elseif ( $code < 0xd30c ) {
3031 return "\xe3\x85\x8c";
3032 } elseif ( $code < 0xd558 ) {
3033 return "\xe3\x85\x8d";
3035 return "\xe3\x85\x8e";
3042 function initEncoding() {
3043 # Some languages may have an alternate char encoding option
3044 # (Esperanto X-coding, Japanese furigana conversion, etc)
3045 # If this language is used as the primary content language,
3046 # an override to the defaults can be set here on startup.
3053 function recodeForEdit( $s ) {
3054 # For some languages we'll want to explicitly specify
3055 # which characters make it into the edit box raw
3056 # or are converted in some way or another.
3057 global $wgEditEncoding;
3058 if ( $wgEditEncoding == '' ||
$wgEditEncoding == 'UTF-8' ) {
3061 return $this->iconv( 'UTF-8', $wgEditEncoding, $s );
3069 function recodeInput( $s ) {
3070 # Take the previous into account.
3071 global $wgEditEncoding;
3072 if ( $wgEditEncoding != '' ) {
3073 $enc = $wgEditEncoding;
3077 if ( $enc == 'UTF-8' ) {
3080 return $this->iconv( $enc, 'UTF-8', $s );
3085 * Convert a UTF-8 string to normal form C. In Malayalam and Arabic, this
3086 * also cleans up certain backwards-compatible sequences, converting them
3087 * to the modern Unicode equivalent.
3089 * This is language-specific for performance reasons only.
3095 function normalize( $s ) {
3096 global $wgAllUnicodeFixes;
3097 $s = UtfNormal\Validator
::cleanUp( $s );
3098 if ( $wgAllUnicodeFixes ) {
3099 $s = $this->transformUsingPairFile( 'normalize-ar.ser', $s );
3100 $s = $this->transformUsingPairFile( 'normalize-ml.ser', $s );
3107 * Transform a string using serialized data stored in the given file (which
3108 * must be in the serialized subdirectory of $IP). The file contains pairs
3109 * mapping source characters to destination characters.
3111 * The data is cached in process memory. This will go faster if you have the
3112 * FastStringSearch extension.
3114 * @param string $file
3115 * @param string $string
3117 * @throws MWException
3120 function transformUsingPairFile( $file, $string ) {
3121 if ( !isset( $this->transformData
[$file] ) ) {
3122 $data = wfGetPrecompiledData( $file );
3123 if ( $data === false ) {
3124 throw new MWException( __METHOD__
. ": The transformation file $file is missing" );
3126 $this->transformData
[$file] = new ReplacementArray( $data );
3128 return $this->transformData
[$file]->replace( $string );
3132 * For right-to-left language support
3137 return self
::$dataCache->getItem( $this->mCode
, 'rtl' );
3141 * Return the correct HTML 'dir' attribute value for this language.
3145 return $this->isRTL() ?
'rtl' : 'ltr';
3149 * Return 'left' or 'right' as appropriate alignment for line-start
3150 * for this language's text direction.
3152 * Should be equivalent to CSS3 'start' text-align value....
3156 function alignStart() {
3157 return $this->isRTL() ?
'right' : 'left';
3161 * Return 'right' or 'left' as appropriate alignment for line-end
3162 * for this language's text direction.
3164 * Should be equivalent to CSS3 'end' text-align value....
3168 function alignEnd() {
3169 return $this->isRTL() ?
'left' : 'right';
3173 * A hidden direction mark (LRM or RLM), depending on the language direction.
3174 * Unlike getDirMark(), this function returns the character as an HTML entity.
3175 * This function should be used when the output is guaranteed to be HTML,
3176 * because it makes the output HTML source code more readable. When
3177 * the output is plain text or can be escaped, getDirMark() should be used.
3179 * @param bool $opposite Get the direction mark opposite to your language
3183 function getDirMarkEntity( $opposite = false ) {
3185 return $this->isRTL() ?
'‎' : '‏';
3187 return $this->isRTL() ?
'‏' : '‎';
3191 * A hidden direction mark (LRM or RLM), depending on the language direction.
3192 * This function produces them as invisible Unicode characters and
3193 * the output may be hard to read and debug, so it should only be used
3194 * when the output is plain text or can be escaped. When the output is
3195 * HTML, use getDirMarkEntity() instead.
3197 * @param bool $opposite Get the direction mark opposite to your language
3200 function getDirMark( $opposite = false ) {
3201 $lrm = "\xE2\x80\x8E"; # LEFT-TO-RIGHT MARK, commonly abbreviated LRM
3202 $rlm = "\xE2\x80\x8F"; # RIGHT-TO-LEFT MARK, commonly abbreviated RLM
3204 return $this->isRTL() ?
$lrm : $rlm;
3206 return $this->isRTL() ?
$rlm : $lrm;
3212 function capitalizeAllNouns() {
3213 return self
::$dataCache->getItem( $this->mCode
, 'capitalizeAllNouns' );
3217 * An arrow, depending on the language direction.
3219 * @param string $direction The direction of the arrow: forwards (default),
3220 * backwards, left, right, up, down.
3223 function getArrow( $direction = 'forwards' ) {
3224 switch ( $direction ) {
3226 return $this->isRTL() ?
'←' : '→';
3228 return $this->isRTL() ?
'→' : '←';
3241 * To allow "foo[[bar]]" to extend the link over the whole word "foobar"
3245 function linkPrefixExtension() {
3246 return self
::$dataCache->getItem( $this->mCode
, 'linkPrefixExtension' );
3250 * Get all magic words from cache.
3253 function getMagicWords() {
3254 return self
::$dataCache->getItem( $this->mCode
, 'magicWords' );
3258 * Run the LanguageGetMagic hook once.
3260 protected function doMagicHook() {
3261 if ( $this->mMagicHookDone
) {
3264 $this->mMagicHookDone
= true;
3265 Hooks
::run( 'LanguageGetMagic', array( &$this->mMagicExtensions
, $this->getCode() ) );
3269 * Fill a MagicWord object with data from here
3271 * @param MagicWord $mw
3273 function getMagic( $mw ) {
3274 // Saves a function call
3275 if ( !$this->mMagicHookDone
) {
3276 $this->doMagicHook();
3279 if ( isset( $this->mMagicExtensions
[$mw->mId
] ) ) {
3280 $rawEntry = $this->mMagicExtensions
[$mw->mId
];
3282 $rawEntry = self
::$dataCache->getSubitem(
3283 $this->mCode
, 'magicWords', $mw->mId
);
3286 if ( !is_array( $rawEntry ) ) {
3287 wfWarn( "\"$rawEntry\" is not a valid magic word for \"$mw->mId\"" );
3289 $mw->mCaseSensitive
= $rawEntry[0];
3290 $mw->mSynonyms
= array_slice( $rawEntry, 1 );
3295 * Add magic words to the extension array
3297 * @param array $newWords
3299 function addMagicWordsByLang( $newWords ) {
3300 $fallbackChain = $this->getFallbackLanguages();
3301 $fallbackChain = array_reverse( $fallbackChain );
3302 foreach ( $fallbackChain as $code ) {
3303 if ( isset( $newWords[$code] ) ) {
3304 $this->mMagicExtensions
= $newWords[$code] +
$this->mMagicExtensions
;
3310 * Get special page names, as an associative array
3311 * canonical name => array of valid names, including aliases
3314 function getSpecialPageAliases() {
3315 // Cache aliases because it may be slow to load them
3316 if ( is_null( $this->mExtendedSpecialPageAliases
) ) {
3318 $this->mExtendedSpecialPageAliases
=
3319 self
::$dataCache->getItem( $this->mCode
, 'specialPageAliases' );
3320 Hooks
::run( 'LanguageGetSpecialPageAliases',
3321 array( &$this->mExtendedSpecialPageAliases
, $this->getCode() ) );
3324 return $this->mExtendedSpecialPageAliases
;
3328 * Italic is unsuitable for some languages
3330 * @param string $text The text to be emphasized.
3333 function emphasize( $text ) {
3334 return "<em>$text</em>";
3338 * Normally we output all numbers in plain en_US style, that is
3339 * 293,291.235 for twohundredninetythreethousand-twohundredninetyone
3340 * point twohundredthirtyfive. However this is not suitable for all
3341 * languages, some such as Punjabi want ੨੯੩,੨੯੫.੨੩੫ and others such as
3342 * Icelandic just want to use commas instead of dots, and dots instead
3343 * of commas like "293.291,235".
3345 * An example of this function being called:
3347 * wfMessage( 'message' )->numParams( $num )->text()
3350 * See $separatorTransformTable on MessageIs.php for
3351 * the , => . and . => , implementation.
3353 * @todo check if it's viable to use localeconv() for the decimal separator thing.
3354 * @param int|float $number The string to be formatted, should be an integer
3355 * or a floating point number.
3356 * @param bool $nocommafy Set to true for special numbers like dates
3359 public function formatNum( $number, $nocommafy = false ) {
3360 global $wgTranslateNumerals;
3361 if ( !$nocommafy ) {
3362 $number = $this->commafy( $number );
3363 $s = $this->separatorTransformTable();
3365 $number = strtr( $number, $s );
3369 if ( $wgTranslateNumerals ) {
3370 $s = $this->digitTransformTable();
3372 $number = strtr( $number, $s );
3380 * Front-end for non-commafied formatNum
3382 * @param int|float $number The string to be formatted, should be an integer
3383 * or a floating point number.
3387 public function formatNumNoSeparators( $number ) {
3388 return $this->formatNum( $number, true );
3392 * @param string $number
3395 public function parseFormattedNumber( $number ) {
3396 $s = $this->digitTransformTable();
3398 // eliminate empty array values such as ''. (bug 64347)
3399 $s = array_filter( $s );
3400 $number = strtr( $number, array_flip( $s ) );
3403 $s = $this->separatorTransformTable();
3405 // eliminate empty array values such as ''. (bug 64347)
3406 $s = array_filter( $s );
3407 $number = strtr( $number, array_flip( $s ) );
3410 $number = strtr( $number, array( ',' => '' ) );
3415 * Adds commas to a given number
3417 * @param mixed $number
3420 function commafy( $number ) {
3421 $digitGroupingPattern = $this->digitGroupingPattern();
3422 if ( $number === null ) {
3426 if ( !$digitGroupingPattern ||
$digitGroupingPattern === "###,###,###" ) {
3427 // default grouping is at thousands, use the same for ###,###,### pattern too.
3428 return strrev( (string)preg_replace( '/(\d{3})(?=\d)(?!\d*\.)/', '$1,', strrev( $number ) ) );
3430 // Ref: http://cldr.unicode.org/translation/number-patterns
3432 if ( intval( $number ) < 0 ) {
3433 // For negative numbers apply the algorithm like positive number and add sign.
3435 $number = substr( $number, 1 );
3437 $integerPart = array();
3438 $decimalPart = array();
3439 $numMatches = preg_match_all( "/(#+)/", $digitGroupingPattern, $matches );
3440 preg_match( "/\d+/", $number, $integerPart );
3441 preg_match( "/\.\d*/", $number, $decimalPart );
3442 $groupedNumber = ( count( $decimalPart ) > 0 ) ?
$decimalPart[0] : "";
3443 if ( $groupedNumber === $number ) {
3444 // the string does not have any number part. Eg: .12345
3445 return $sign . $groupedNumber;
3447 $start = $end = ($integerPart) ?
strlen( $integerPart[0] ) : 0;
3448 while ( $start > 0 ) {
3449 $match = $matches[0][$numMatches - 1];
3450 $matchLen = strlen( $match );
3451 $start = $end - $matchLen;
3455 $groupedNumber = substr( $number, $start, $end -$start ) . $groupedNumber;
3457 if ( $numMatches > 1 ) {
3458 // use the last pattern for the rest of the number
3462 $groupedNumber = "," . $groupedNumber;
3465 return $sign . $groupedNumber;
3472 function digitGroupingPattern() {
3473 return self
::$dataCache->getItem( $this->mCode
, 'digitGroupingPattern' );
3479 function digitTransformTable() {
3480 return self
::$dataCache->getItem( $this->mCode
, 'digitTransformTable' );
3486 function separatorTransformTable() {
3487 return self
::$dataCache->getItem( $this->mCode
, 'separatorTransformTable' );
3491 * Take a list of strings and build a locale-friendly comma-separated
3492 * list, using the local comma-separator message.
3493 * The last two strings are chained with an "and".
3494 * NOTE: This function will only work with standard numeric array keys (0, 1, 2…)
3496 * @param string[] $l
3499 function listToText( array $l ) {
3500 $m = count( $l ) - 1;
3505 $and = $this->msg( 'and' )->escaped();
3506 $space = $this->msg( 'word-separator' )->escaped();
3508 $comma = $this->msg( 'comma-separator' )->escaped();
3512 for ( $i = $m - 1; $i >= 0; $i-- ) {
3513 if ( $i == $m - 1 ) {
3514 $s = $l[$i] . $and . $space . $s;
3516 $s = $l[$i] . $comma . $s;
3523 * Take a list of strings and build a locale-friendly comma-separated
3524 * list, using the local comma-separator message.
3525 * @param string[] $list Array of strings to put in a comma list
3528 function commaList( array $list ) {
3530 wfMessage( 'comma-separator' )->inLanguage( $this )->escaped(),
3536 * Take a list of strings and build a locale-friendly semicolon-separated
3537 * list, using the local semicolon-separator message.
3538 * @param string[] $list Array of strings to put in a semicolon list
3541 function semicolonList( array $list ) {
3543 wfMessage( 'semicolon-separator' )->inLanguage( $this )->escaped(),
3549 * Same as commaList, but separate it with the pipe instead.
3550 * @param string[] $list Array of strings to put in a pipe list
3553 function pipeList( array $list ) {
3555 wfMessage( 'pipe-separator' )->inLanguage( $this )->escaped(),
3561 * Truncate a string to a specified length in bytes, appending an optional
3562 * string (e.g. for ellipses)
3564 * The database offers limited byte lengths for some columns in the database;
3565 * multi-byte character sets mean we need to ensure that only whole characters
3566 * are included, otherwise broken characters can be passed to the user
3568 * If $length is negative, the string will be truncated from the beginning
3570 * @param string $string String to truncate
3571 * @param int $length Maximum length (including ellipses)
3572 * @param string $ellipsis String to append to the truncated text
3573 * @param bool $adjustLength Subtract length of ellipsis from $length.
3574 * $adjustLength was introduced in 1.18, before that behaved as if false.
3577 function truncate( $string, $length, $ellipsis = '...', $adjustLength = true ) {
3578 # Use the localized ellipsis character
3579 if ( $ellipsis == '...' ) {
3580 $ellipsis = wfMessage( 'ellipsis' )->inLanguage( $this )->escaped();
3582 # Check if there is no need to truncate
3583 if ( $length == 0 ) {
3584 return $ellipsis; // convention
3585 } elseif ( strlen( $string ) <= abs( $length ) ) {
3586 return $string; // no need to truncate
3588 $stringOriginal = $string;
3589 # If ellipsis length is >= $length then we can't apply $adjustLength
3590 if ( $adjustLength && strlen( $ellipsis ) >= abs( $length ) ) {
3591 $string = $ellipsis; // this can be slightly unexpected
3592 # Otherwise, truncate and add ellipsis...
3594 $eLength = $adjustLength ?
strlen( $ellipsis ) : 0;
3595 if ( $length > 0 ) {
3596 $length -= $eLength;
3597 $string = substr( $string, 0, $length ); // xyz...
3598 $string = $this->removeBadCharLast( $string );
3599 $string = rtrim( $string );
3600 $string = $string . $ellipsis;
3602 $length +
= $eLength;
3603 $string = substr( $string, $length ); // ...xyz
3604 $string = $this->removeBadCharFirst( $string );
3605 $string = ltrim( $string );
3606 $string = $ellipsis . $string;
3609 # Do not truncate if the ellipsis makes the string longer/equal (bug 22181).
3610 # This check is *not* redundant if $adjustLength, due to the single case where
3611 # LEN($ellipsis) > ABS($limit arg); $stringOriginal could be shorter than $string.
3612 if ( strlen( $string ) < strlen( $stringOriginal ) ) {
3615 return $stringOriginal;
3620 * Remove bytes that represent an incomplete Unicode character
3621 * at the end of string (e.g. bytes of the char are missing)
3623 * @param string $string
3626 protected function removeBadCharLast( $string ) {
3627 if ( $string != '' ) {
3628 $char = ord( $string[strlen( $string ) - 1] );
3630 if ( $char >= 0xc0 ) {
3631 # We got the first byte only of a multibyte char; remove it.
3632 $string = substr( $string, 0, -1 );
3633 } elseif ( $char >= 0x80 &&
3634 preg_match( '/^(.*)(?:[\xe0-\xef][\x80-\xbf]|' .
3635 '[\xf0-\xf7][\x80-\xbf]{1,2})$/', $string, $m )
3637 # We chopped in the middle of a character; remove it
3645 * Remove bytes that represent an incomplete Unicode character
3646 * at the start of string (e.g. bytes of the char are missing)
3648 * @param string $string
3651 protected function removeBadCharFirst( $string ) {
3652 if ( $string != '' ) {
3653 $char = ord( $string[0] );
3654 if ( $char >= 0x80 && $char < 0xc0 ) {
3655 # We chopped in the middle of a character; remove the whole thing
3656 $string = preg_replace( '/^[\x80-\xbf]+/', '', $string );
3663 * Truncate a string of valid HTML to a specified length in bytes,
3664 * appending an optional string (e.g. for ellipses), and return valid HTML
3666 * This is only intended for styled/linked text, such as HTML with
3667 * tags like <span> and <a>, were the tags are self-contained (valid HTML).
3668 * Also, this will not detect things like "display:none" CSS.
3670 * Note: since 1.18 you do not need to leave extra room in $length for ellipses.
3672 * @param string $text HTML string to truncate
3673 * @param int $length (zero/positive) Maximum length (including ellipses)
3674 * @param string $ellipsis String to append to the truncated text
3677 function truncateHtml( $text, $length, $ellipsis = '...' ) {
3678 # Use the localized ellipsis character
3679 if ( $ellipsis == '...' ) {
3680 $ellipsis = wfMessage( 'ellipsis' )->inLanguage( $this )->escaped();
3682 # Check if there is clearly no need to truncate
3683 if ( $length <= 0 ) {
3684 return $ellipsis; // no text shown, nothing to format (convention)
3685 } elseif ( strlen( $text ) <= $length ) {
3686 return $text; // string short enough even *with* HTML (short-circuit)
3689 $dispLen = 0; // innerHTML legth so far
3690 $testingEllipsis = false; // checking if ellipses will make string longer/equal?
3691 $tagType = 0; // 0-open, 1-close
3692 $bracketState = 0; // 1-tag start, 2-tag name, 0-neither
3693 $entityState = 0; // 0-not entity, 1-entity
3694 $tag = $ret = ''; // accumulated tag name, accumulated result string
3695 $openTags = array(); // open tag stack
3696 $maybeState = null; // possible truncation state
3698 $textLen = strlen( $text );
3699 $neLength = max( 0, $length - strlen( $ellipsis ) ); // non-ellipsis len if truncated
3700 for ( $pos = 0; true; ++
$pos ) {
3701 # Consider truncation once the display length has reached the maximim.
3702 # We check if $dispLen > 0 to grab tags for the $neLength = 0 case.
3703 # Check that we're not in the middle of a bracket/entity...
3704 if ( $dispLen && $dispLen >= $neLength && $bracketState == 0 && !$entityState ) {
3705 if ( !$testingEllipsis ) {
3706 $testingEllipsis = true;
3707 # Save where we are; we will truncate here unless there turn out to
3708 # be so few remaining characters that truncation is not necessary.
3709 if ( !$maybeState ) { // already saved? ($neLength = 0 case)
3710 $maybeState = array( $ret, $openTags ); // save state
3712 } elseif ( $dispLen > $length && $dispLen > strlen( $ellipsis ) ) {
3713 # String in fact does need truncation, the truncation point was OK.
3714 list( $ret, $openTags ) = $maybeState; // reload state
3715 $ret = $this->removeBadCharLast( $ret ); // multi-byte char fix
3716 $ret .= $ellipsis; // add ellipsis
3720 if ( $pos >= $textLen ) {
3721 break; // extra iteration just for above checks
3724 # Read the next char...
3726 $lastCh = $pos ?
$text[$pos - 1] : '';
3727 $ret .= $ch; // add to result string
3729 $this->truncate_endBracket( $tag, $tagType, $lastCh, $openTags ); // for bad HTML
3730 $entityState = 0; // for bad HTML
3731 $bracketState = 1; // tag started (checking for backslash)
3732 } elseif ( $ch == '>' ) {
3733 $this->truncate_endBracket( $tag, $tagType, $lastCh, $openTags );
3734 $entityState = 0; // for bad HTML
3735 $bracketState = 0; // out of brackets
3736 } elseif ( $bracketState == 1 ) {
3738 $tagType = 1; // close tag (e.g. "</span>")
3740 $tagType = 0; // open tag (e.g. "<span>")
3743 $bracketState = 2; // building tag name
3744 } elseif ( $bracketState == 2 ) {
3748 // Name found (e.g. "<a href=..."), add on tag attributes...
3749 $pos +
= $this->truncate_skip( $ret, $text, "<>", $pos +
1 );
3751 } elseif ( $bracketState == 0 ) {
3752 if ( $entityState ) {
3755 $dispLen++
; // entity is one displayed char
3758 if ( $neLength == 0 && !$maybeState ) {
3759 // Save state without $ch. We want to *hit* the first
3760 // display char (to get tags) but not *use* it if truncating.
3761 $maybeState = array( substr( $ret, 0, -1 ), $openTags );
3764 $entityState = 1; // entity found, (e.g. " ")
3766 $dispLen++
; // this char is displayed
3767 // Add the next $max display text chars after this in one swoop...
3768 $max = ( $testingEllipsis ?
$length : $neLength ) - $dispLen;
3769 $skipped = $this->truncate_skip( $ret, $text, "<>&", $pos +
1, $max );
3770 $dispLen +
= $skipped;
3776 // Close the last tag if left unclosed by bad HTML
3777 $this->truncate_endBracket( $tag, $text[$textLen - 1], $tagType, $openTags );
3778 while ( count( $openTags ) > 0 ) {
3779 $ret .= '</' . array_pop( $openTags ) . '>'; // close open tags
3785 * truncateHtml() helper function
3786 * like strcspn() but adds the skipped chars to $ret
3788 * @param string $ret
3789 * @param string $text
3790 * @param string $search
3792 * @param null|int $len
3795 private function truncate_skip( &$ret, $text, $search, $start, $len = null ) {
3796 if ( $len === null ) {
3797 $len = -1; // -1 means "no limit" for strcspn
3798 } elseif ( $len < 0 ) {
3802 if ( $start < strlen( $text ) ) {
3803 $skipCount = strcspn( $text, $search, $start, $len );
3804 $ret .= substr( $text, $start, $skipCount );
3810 * truncateHtml() helper function
3811 * (a) push or pop $tag from $openTags as needed
3812 * (b) clear $tag value
3813 * @param string &$tag Current HTML tag name we are looking at
3814 * @param int $tagType (0-open tag, 1-close tag)
3815 * @param string $lastCh Character before the '>' that ended this tag
3816 * @param array &$openTags Open tag stack (not accounting for $tag)
3818 private function truncate_endBracket( &$tag, $tagType, $lastCh, &$openTags ) {
3819 $tag = ltrim( $tag );
3821 if ( $tagType == 0 && $lastCh != '/' ) {
3822 $openTags[] = $tag; // tag opened (didn't close itself)
3823 } elseif ( $tagType == 1 ) {
3824 if ( $openTags && $tag == $openTags[count( $openTags ) - 1] ) {
3825 array_pop( $openTags ); // tag closed
3833 * Grammatical transformations, needed for inflected languages
3834 * Invoked by putting {{grammar:case|word}} in a message
3836 * @param string $word
3837 * @param string $case
3840 function convertGrammar( $word, $case ) {
3841 global $wgGrammarForms;
3842 if ( isset( $wgGrammarForms[$this->getCode()][$case][$word] ) ) {
3843 return $wgGrammarForms[$this->getCode()][$case][$word];
3849 * Get the grammar forms for the content language
3850 * @return array Array of grammar forms
3853 function getGrammarForms() {
3854 global $wgGrammarForms;
3855 if ( isset( $wgGrammarForms[$this->getCode()] )
3856 && is_array( $wgGrammarForms[$this->getCode()] )
3858 return $wgGrammarForms[$this->getCode()];
3864 * Provides an alternative text depending on specified gender.
3865 * Usage {{gender:username|masculine|feminine|unknown}}.
3866 * username is optional, in which case the gender of current user is used,
3867 * but only in (some) interface messages; otherwise default gender is used.
3869 * If no forms are given, an empty string is returned. If only one form is
3870 * given, it will be returned unconditionally. These details are implied by
3871 * the caller and cannot be overridden in subclasses.
3873 * If three forms are given, the default is to use the third (unknown) form.
3874 * If fewer than three forms are given, the default is to use the first (masculine) form.
3875 * These details can be overridden in subclasses.
3877 * @param string $gender
3878 * @param array $forms
3882 function gender( $gender, $forms ) {
3883 if ( !count( $forms ) ) {
3886 $forms = $this->preConvertPlural( $forms, 2 );
3887 if ( $gender === 'male' ) {
3890 if ( $gender === 'female' ) {
3893 return isset( $forms[2] ) ?
$forms[2] : $forms[0];
3897 * Plural form transformations, needed for some languages.
3898 * For example, there are 3 form of plural in Russian and Polish,
3899 * depending on "count mod 10". See [[w:Plural]]
3900 * For English it is pretty simple.
3902 * Invoked by putting {{plural:count|wordform1|wordform2}}
3903 * or {{plural:count|wordform1|wordform2|wordform3}}
3905 * Example: {{plural:{{NUMBEROFARTICLES}}|article|articles}}
3907 * @param int $count Non-localized number
3908 * @param array $forms Different plural forms
3909 * @return string Correct form of plural for $count in this language
3911 function convertPlural( $count, $forms ) {
3912 // Handle explicit n=pluralform cases
3913 $forms = $this->handleExplicitPluralForms( $count, $forms );
3914 if ( is_string( $forms ) ) {
3917 if ( !count( $forms ) ) {
3921 $pluralForm = $this->getPluralRuleIndexNumber( $count );
3922 $pluralForm = min( $pluralForm, count( $forms ) - 1 );
3923 return $forms[$pluralForm];
3927 * Handles explicit plural forms for Language::convertPlural()
3929 * In {{PLURAL:$1|0=nothing|one|many}}, 0=nothing will be returned if $1 equals zero.
3930 * If an explicitly defined plural form matches the $count, then
3931 * string value returned, otherwise array returned for further consideration
3932 * by CLDR rules or overridden convertPlural().
3936 * @param int $count Non-localized number
3937 * @param array $forms Different plural forms
3939 * @return array|string
3941 protected function handleExplicitPluralForms( $count, array $forms ) {
3942 foreach ( $forms as $index => $form ) {
3943 if ( preg_match( '/\d+=/i', $form ) ) {
3944 $pos = strpos( $form, '=' );
3945 if ( substr( $form, 0, $pos ) === (string)$count ) {
3946 return substr( $form, $pos +
1 );
3948 unset( $forms[$index] );
3951 return array_values( $forms );
3955 * Checks that convertPlural was given an array and pads it to requested
3956 * amount of forms by copying the last one.
3958 * @param array $forms Array of forms given to convertPlural
3959 * @param int $count How many forms should there be at least
3960 * @return array Padded array of forms or an exception if not an array
3962 protected function preConvertPlural( /* Array */ $forms, $count ) {
3963 while ( count( $forms ) < $count ) {
3964 $forms[] = $forms[count( $forms ) - 1];
3970 * @todo Maybe translate block durations. Note that this function is somewhat misnamed: it
3971 * deals with translating the *duration* ("1 week", "4 days", etc), not the expiry time
3972 * (which is an absolute timestamp). Please note: do NOT add this blindly, as it is used
3973 * on old expiry lengths recorded in log entries. You'd need to provide the start date to
3976 * @param string $str The validated block duration in English
3977 * @return string Somehow translated block duration
3978 * @see LanguageFi.php for example implementation
3980 function translateBlockExpiry( $str ) {
3981 $duration = SpecialBlock
::getSuggestedDurations( $this );
3982 foreach ( $duration as $show => $value ) {
3983 if ( strcmp( $str, $value ) == 0 ) {
3984 return htmlspecialchars( trim( $show ) );
3988 if ( wfIsInfinity( $str ) ) {
3989 foreach ( $duration as $show => $value ) {
3990 if ( wfIsInfinity( $value ) ) {
3991 return htmlspecialchars( trim( $show ) );
3996 // If all else fails, return a standard duration or timestamp description.
3997 $time = strtotime( $str, 0 );
3998 if ( $time === false ) { // Unknown format. Return it as-is in case.
4000 } elseif ( $time !== strtotime( $str, 1 ) ) { // It's a relative timestamp.
4001 // $time is relative to 0 so it's a duration length.
4002 return $this->formatDuration( $time );
4003 } else { // It's an absolute timestamp.
4004 if ( $time === 0 ) {
4005 // wfTimestamp() handles 0 as current time instead of epoch.
4006 return $this->timeanddate( '19700101000000' );
4008 return $this->timeanddate( $time );
4014 * languages like Chinese need to be segmented in order for the diff
4017 * @param string $text
4020 public function segmentForDiff( $text ) {
4025 * and unsegment to show the result
4027 * @param string $text
4030 public function unsegmentForDiff( $text ) {
4035 * Return the LanguageConverter used in the Language
4038 * @return LanguageConverter
4040 public function getConverter() {
4041 return $this->mConverter
;
4045 * convert text to all supported variants
4047 * @param string $text
4050 public function autoConvertToAllVariants( $text ) {
4051 return $this->mConverter
->autoConvertToAllVariants( $text );
4055 * convert text to different variants of a language.
4057 * @param string $text
4060 public function convert( $text ) {
4061 return $this->mConverter
->convert( $text );
4065 * Convert a Title object to a string in the preferred variant
4067 * @param Title $title
4070 public function convertTitle( $title ) {
4071 return $this->mConverter
->convertTitle( $title );
4075 * Convert a namespace index to a string in the preferred variant
4080 public function convertNamespace( $ns ) {
4081 return $this->mConverter
->convertNamespace( $ns );
4085 * Check if this is a language with variants
4089 public function hasVariants() {
4090 return count( $this->getVariants() ) > 1;
4094 * Check if the language has the specific variant
4097 * @param string $variant
4100 public function hasVariant( $variant ) {
4101 return (bool)$this->mConverter
->validateVariant( $variant );
4105 * Put custom tags (e.g. -{ }-) around math to prevent conversion
4107 * @param string $text
4109 * @deprecated since 1.22 is no longer used
4111 public function armourMath( $text ) {
4112 return $this->mConverter
->armourMath( $text );
4116 * Perform output conversion on a string, and encode for safe HTML output.
4117 * @param string $text Text to be converted
4118 * @param bool $isTitle Whether this conversion is for the article title
4120 * @todo this should get integrated somewhere sane
4122 public function convertHtml( $text, $isTitle = false ) {
4123 return htmlspecialchars( $this->convert( $text, $isTitle ) );
4127 * @param string $key
4130 public function convertCategoryKey( $key ) {
4131 return $this->mConverter
->convertCategoryKey( $key );
4135 * Get the list of variants supported by this language
4136 * see sample implementation in LanguageZh.php
4138 * @return array An array of language codes
4140 public function getVariants() {
4141 return $this->mConverter
->getVariants();
4147 public function getPreferredVariant() {
4148 return $this->mConverter
->getPreferredVariant();
4154 public function getDefaultVariant() {
4155 return $this->mConverter
->getDefaultVariant();
4161 public function getURLVariant() {
4162 return $this->mConverter
->getURLVariant();
4166 * If a language supports multiple variants, it is
4167 * possible that non-existing link in one variant
4168 * actually exists in another variant. this function
4169 * tries to find it. See e.g. LanguageZh.php
4170 * The input parameters may be modified upon return
4172 * @param string &$link The name of the link
4173 * @param Title &$nt The title object of the link
4174 * @param bool $ignoreOtherCond To disable other conditions when
4175 * we need to transclude a template or update a category's link
4177 public function findVariantLink( &$link, &$nt, $ignoreOtherCond = false ) {
4178 $this->mConverter
->findVariantLink( $link, $nt, $ignoreOtherCond );
4182 * returns language specific options used by User::getPageRenderHash()
4183 * for example, the preferred language variant
4187 function getExtraHashOptions() {
4188 return $this->mConverter
->getExtraHashOptions();
4192 * For languages that support multiple variants, the title of an
4193 * article may be displayed differently in different variants. this
4194 * function returns the apporiate title defined in the body of the article.
4198 public function getParsedTitle() {
4199 return $this->mConverter
->getParsedTitle();
4203 * Prepare external link text for conversion. When the text is
4204 * a URL, it shouldn't be converted, and it'll be wrapped in
4205 * the "raw" tag (-{R| }-) to prevent conversion.
4207 * This function is called "markNoConversion" for historical
4210 * @param string $text Text to be used for external link
4211 * @param bool $noParse Wrap it without confirming it's a real URL first
4212 * @return string The tagged text
4214 public function markNoConversion( $text, $noParse = false ) {
4215 // Excluding protocal-relative URLs may avoid many false positives.
4216 if ( $noParse ||
preg_match( '/^(?:' . wfUrlProtocolsWithoutProtRel() . ')/', $text ) ) {
4217 return $this->mConverter
->markNoConversion( $text );
4224 * A regular expression to match legal word-trailing characters
4225 * which should be merged onto a link of the form [[foo]]bar.
4229 public function linkTrail() {
4230 return self
::$dataCache->getItem( $this->mCode
, 'linkTrail' );
4234 * A regular expression character set to match legal word-prefixing
4235 * characters which should be merged onto a link of the form foo[[bar]].
4239 public function linkPrefixCharset() {
4240 return self
::$dataCache->getItem( $this->mCode
, 'linkPrefixCharset' );
4244 * @deprecated since 1.24, will be removed in 1.25
4247 function getLangObj() {
4248 wfDeprecated( __METHOD__
, '1.24' );
4253 * Get the "parent" language which has a converter to convert a "compatible" language
4254 * (in another variant) to this language (eg. zh for zh-cn, but not en for en-gb).
4256 * @return Language|null
4259 public function getParentLanguage() {
4260 if ( $this->mParentLanguage
!== false ) {
4261 return $this->mParentLanguage
;
4264 $pieces = explode( '-', $this->getCode() );
4266 if ( !in_array( $code, LanguageConverter
::$languagesWithVariants ) ) {
4267 $this->mParentLanguage
= null;
4270 $lang = Language
::factory( $code );
4271 if ( !$lang->hasVariant( $this->getCode() ) ) {
4272 $this->mParentLanguage
= null;
4276 $this->mParentLanguage
= $lang;
4281 * Get the RFC 3066 code for this language object
4283 * NOTE: The return value of this function is NOT HTML-safe and must be escaped with
4284 * htmlspecialchars() or similar
4288 public function getCode() {
4289 return $this->mCode
;
4293 * Get the code in Bcp47 format which we can use
4294 * inside of html lang="" tags.
4296 * NOTE: The return value of this function is NOT HTML-safe and must be escaped with
4297 * htmlspecialchars() or similar.
4302 public function getHtmlCode() {
4303 if ( is_null( $this->mHtmlCode
) ) {
4304 $this->mHtmlCode
= wfBCP47( $this->getCode() );
4306 return $this->mHtmlCode
;
4310 * @param string $code
4312 public function setCode( $code ) {
4313 $this->mCode
= $code;
4314 // Ensure we don't leave incorrect cached data lying around
4315 $this->mHtmlCode
= null;
4316 $this->mParentLanguage
= false;
4320 * Get the name of a file for a certain language code
4321 * @param string $prefix Prepend this to the filename
4322 * @param string $code Language code
4323 * @param string $suffix Append this to the filename
4324 * @throws MWException
4325 * @return string $prefix . $mangledCode . $suffix
4327 public static function getFileName( $prefix = 'Language', $code, $suffix = '.php' ) {
4328 if ( !self
::isValidBuiltInCode( $code ) ) {
4329 throw new MWException( "Invalid language code \"$code\"" );
4332 return $prefix . str_replace( '-', '_', ucfirst( $code ) ) . $suffix;
4336 * Get the language code from a file name. Inverse of getFileName()
4337 * @param string $filename $prefix . $languageCode . $suffix
4338 * @param string $prefix Prefix before the language code
4339 * @param string $suffix Suffix after the language code
4340 * @return string Language code, or false if $prefix or $suffix isn't found
4342 public static function getCodeFromFileName( $filename, $prefix = 'Language', $suffix = '.php' ) {
4344 preg_match( '/' . preg_quote( $prefix, '/' ) . '([A-Z][a-z_]+)' .
4345 preg_quote( $suffix, '/' ) . '/', $filename, $m );
4346 if ( !count( $m ) ) {
4349 return str_replace( '_', '-', strtolower( $m[1] ) );
4353 * @param string $code
4356 public static function getMessagesFileName( $code ) {
4358 $file = self
::getFileName( "$IP/languages/messages/Messages", $code, '.php' );
4359 Hooks
::run( 'Language::getMessagesFileName', array( $code, &$file ) );
4364 * @param string $code
4368 public static function getJsonMessagesFileName( $code ) {
4371 if ( !self
::isValidBuiltInCode( $code ) ) {
4372 throw new MWException( "Invalid language code \"$code\"" );
4375 return "$IP/languages/i18n/$code.json";
4379 * @param string $code
4382 public static function getClassFileName( $code ) {
4384 return self
::getFileName( "$IP/languages/classes/Language", $code, '.php' );
4388 * Get the first fallback for a given language.
4390 * @param string $code
4392 * @return bool|string
4394 public static function getFallbackFor( $code ) {
4395 if ( $code === 'en' ||
!Language
::isValidBuiltInCode( $code ) ) {
4398 $fallbacks = self
::getFallbacksFor( $code );
4399 $first = array_shift( $fallbacks );
4405 * Get the ordered list of fallback languages.
4408 * @param string $code Language code
4411 public static function getFallbacksFor( $code ) {
4412 if ( $code === 'en' ||
!Language
::isValidBuiltInCode( $code ) ) {
4415 $v = self
::getLocalisationCache()->getItem( $code, 'fallback' );
4416 $v = array_map( 'trim', explode( ',', $v ) );
4417 if ( $v[count( $v ) - 1] !== 'en' ) {
4425 * Get the ordered list of fallback languages, ending with the fallback
4426 * language chain for the site language.
4429 * @param string $code Language code
4430 * @return array Array( fallbacks, site fallbacks )
4432 public static function getFallbacksIncludingSiteLanguage( $code ) {
4433 global $wgLanguageCode;
4435 // Usually, we will only store a tiny number of fallback chains, so we
4436 // keep them in static memory.
4437 $cacheKey = "{$code}-{$wgLanguageCode}";
4439 if ( !array_key_exists( $cacheKey, self
::$fallbackLanguageCache ) ) {
4440 $fallbacks = self
::getFallbacksFor( $code );
4442 // Append the site's fallback chain, including the site language itself
4443 $siteFallbacks = self
::getFallbacksFor( $wgLanguageCode );
4444 array_unshift( $siteFallbacks, $wgLanguageCode );
4446 // Eliminate any languages already included in the chain
4447 $siteFallbacks = array_diff( $siteFallbacks, $fallbacks );
4449 self
::$fallbackLanguageCache[$cacheKey] = array( $fallbacks, $siteFallbacks );
4451 return self
::$fallbackLanguageCache[$cacheKey];
4455 * Get all messages for a given language
4456 * WARNING: this may take a long time. If you just need all message *keys*
4457 * but need the *contents* of only a few messages, consider using getMessageKeysFor().
4459 * @param string $code
4463 public static function getMessagesFor( $code ) {
4464 return self
::getLocalisationCache()->getItem( $code, 'messages' );
4468 * Get a message for a given language
4470 * @param string $key
4471 * @param string $code
4475 public static function getMessageFor( $key, $code ) {
4476 return self
::getLocalisationCache()->getSubitem( $code, 'messages', $key );
4480 * Get all message keys for a given language. This is a faster alternative to
4481 * array_keys( Language::getMessagesFor( $code ) )
4484 * @param string $code Language code
4485 * @return array Array of message keys (strings)
4487 public static function getMessageKeysFor( $code ) {
4488 return self
::getLocalisationCache()->getSubItemList( $code, 'messages' );
4492 * @param string $talk
4495 function fixVariableInNamespace( $talk ) {
4496 if ( strpos( $talk, '$1' ) === false ) {
4500 global $wgMetaNamespace;
4501 $talk = str_replace( '$1', $wgMetaNamespace, $talk );
4503 # Allow grammar transformations
4504 # Allowing full message-style parsing would make simple requests
4505 # such as action=raw much more expensive than they need to be.
4506 # This will hopefully cover most cases.
4507 $talk = preg_replace_callback( '/{{grammar:(.*?)\|(.*?)}}/i',
4508 array( &$this, 'replaceGrammarInNamespace' ), $talk );
4509 return str_replace( ' ', '_', $talk );
4516 function replaceGrammarInNamespace( $m ) {
4517 return $this->convertGrammar( trim( $m[2] ), trim( $m[1] ) );
4521 * @throws MWException
4524 static function getCaseMaps() {
4525 static $wikiUpperChars, $wikiLowerChars;
4526 if ( isset( $wikiUpperChars ) ) {
4527 return array( $wikiUpperChars, $wikiLowerChars );
4530 $arr = wfGetPrecompiledData( 'Utf8Case.ser' );
4531 if ( $arr === false ) {
4532 throw new MWException(
4533 "Utf8Case.ser is missing, please run \"make\" in the serialized directory\n" );
4535 $wikiUpperChars = $arr['wikiUpperChars'];
4536 $wikiLowerChars = $arr['wikiLowerChars'];
4537 return array( $wikiUpperChars, $wikiLowerChars );
4541 * Decode an expiry (block, protection, etc) which has come from the DB
4543 * @param string $expiry Database expiry String
4544 * @param bool|int $format True to process using language functions, or TS_ constant
4545 * to return the expiry in a given timestamp
4546 * @param string $inifinity If $format is not true, use this string for infinite expiry
4550 public function formatExpiry( $expiry, $format = true, $infinity = 'infinity' ) {
4552 if ( $dbInfinity === null ) {
4553 $dbInfinity = wfGetDB( DB_SLAVE
)->getInfinity();
4556 if ( $expiry == '' ||
$expiry === 'infinity' ||
$expiry == $dbInfinity ) {
4557 return $format === true
4558 ?
$this->getMessageFromDB( 'infiniteblock' )
4561 return $format === true
4562 ?
$this->timeanddate( $expiry, /* User preference timezone */ true )
4563 : wfTimestamp( $format, $expiry );
4569 * @param int|float $seconds
4570 * @param array $format Optional
4571 * If $format['avoid'] === 'avoidseconds': don't mention seconds if $seconds >= 1 hour.
4572 * If $format['avoid'] === 'avoidminutes': don't mention seconds/minutes if $seconds > 48 hours.
4573 * If $format['noabbrevs'] is true: use 'seconds' and friends instead of 'seconds-abbrev'
4575 * For backwards compatibility, $format may also be one of the strings 'avoidseconds'
4576 * or 'avoidminutes'.
4579 function formatTimePeriod( $seconds, $format = array() ) {
4580 if ( !is_array( $format ) ) {
4581 $format = array( 'avoid' => $format ); // For backwards compatibility
4583 if ( !isset( $format['avoid'] ) ) {
4584 $format['avoid'] = false;
4586 if ( !isset( $format['noabbrevs'] ) ) {
4587 $format['noabbrevs'] = false;
4589 $secondsMsg = wfMessage(
4590 $format['noabbrevs'] ?
'seconds' : 'seconds-abbrev' )->inLanguage( $this );
4591 $minutesMsg = wfMessage(
4592 $format['noabbrevs'] ?
'minutes' : 'minutes-abbrev' )->inLanguage( $this );
4593 $hoursMsg = wfMessage(
4594 $format['noabbrevs'] ?
'hours' : 'hours-abbrev' )->inLanguage( $this );
4595 $daysMsg = wfMessage(
4596 $format['noabbrevs'] ?
'days' : 'days-abbrev' )->inLanguage( $this );
4598 if ( round( $seconds * 10 ) < 100 ) {
4599 $s = $this->formatNum( sprintf( "%.1f", round( $seconds * 10 ) / 10 ) );
4600 $s = $secondsMsg->params( $s )->text();
4601 } elseif ( round( $seconds ) < 60 ) {
4602 $s = $this->formatNum( round( $seconds ) );
4603 $s = $secondsMsg->params( $s )->text();
4604 } elseif ( round( $seconds ) < 3600 ) {
4605 $minutes = floor( $seconds / 60 );
4606 $secondsPart = round( fmod( $seconds, 60 ) );
4607 if ( $secondsPart == 60 ) {
4611 $s = $minutesMsg->params( $this->formatNum( $minutes ) )->text();
4613 $s .= $secondsMsg->params( $this->formatNum( $secondsPart ) )->text();
4614 } elseif ( round( $seconds ) <= 2 * 86400 ) {
4615 $hours = floor( $seconds / 3600 );
4616 $minutes = floor( ( $seconds - $hours * 3600 ) / 60 );
4617 $secondsPart = round( $seconds - $hours * 3600 - $minutes * 60 );
4618 if ( $secondsPart == 60 ) {
4622 if ( $minutes == 60 ) {
4626 $s = $hoursMsg->params( $this->formatNum( $hours ) )->text();
4628 $s .= $minutesMsg->params( $this->formatNum( $minutes ) )->text();
4629 if ( !in_array( $format['avoid'], array( 'avoidseconds', 'avoidminutes' ) ) ) {
4630 $s .= ' ' . $secondsMsg->params( $this->formatNum( $secondsPart ) )->text();
4633 $days = floor( $seconds / 86400 );
4634 if ( $format['avoid'] === 'avoidminutes' ) {
4635 $hours = round( ( $seconds - $days * 86400 ) / 3600 );
4636 if ( $hours == 24 ) {
4640 $s = $daysMsg->params( $this->formatNum( $days ) )->text();
4642 $s .= $hoursMsg->params( $this->formatNum( $hours ) )->text();
4643 } elseif ( $format['avoid'] === 'avoidseconds' ) {
4644 $hours = floor( ( $seconds - $days * 86400 ) / 3600 );
4645 $minutes = round( ( $seconds - $days * 86400 - $hours * 3600 ) / 60 );
4646 if ( $minutes == 60 ) {
4650 if ( $hours == 24 ) {
4654 $s = $daysMsg->params( $this->formatNum( $days ) )->text();
4656 $s .= $hoursMsg->params( $this->formatNum( $hours ) )->text();
4658 $s .= $minutesMsg->params( $this->formatNum( $minutes ) )->text();
4660 $s = $daysMsg->params( $this->formatNum( $days ) )->text();
4662 $s .= $this->formatTimePeriod( $seconds - $days * 86400, $format );
4669 * Format a bitrate for output, using an appropriate
4670 * unit (bps, kbps, Mbps, Gbps, Tbps, Pbps, Ebps, Zbps or Ybps) according to
4671 * the magnitude in question.
4673 * This use base 1000. For base 1024 use formatSize(), for another base
4674 * see formatComputingNumbers().
4679 function formatBitrate( $bps ) {
4680 return $this->formatComputingNumbers( $bps, 1000, "bitrate-$1bits" );
4684 * @param int $size Size of the unit
4685 * @param int $boundary Size boundary (1000, or 1024 in most cases)
4686 * @param string $messageKey Message key to be uesd
4689 function formatComputingNumbers( $size, $boundary, $messageKey ) {
4691 return str_replace( '$1', $this->formatNum( $size ),
4692 $this->getMessageFromDB( str_replace( '$1', '', $messageKey ) )
4695 $sizes = array( '', 'kilo', 'mega', 'giga', 'tera', 'peta', 'exa', 'zeta', 'yotta' );
4698 $maxIndex = count( $sizes ) - 1;
4699 while ( $size >= $boundary && $index < $maxIndex ) {
4704 // For small sizes no decimal places necessary
4707 // For MB and bigger two decimal places are smarter
4710 $msg = str_replace( '$1', $sizes[$index], $messageKey );
4712 $size = round( $size, $round );
4713 $text = $this->getMessageFromDB( $msg );
4714 return str_replace( '$1', $this->formatNum( $size ), $text );
4718 * Format a size in bytes for output, using an appropriate
4719 * unit (B, KB, MB, GB, TB, PB, EB, ZB or YB) according to the magnitude in question
4721 * This method use base 1024. For base 1000 use formatBitrate(), for
4722 * another base see formatComputingNumbers()
4724 * @param int $size Size to format
4725 * @return string Plain text (not HTML)
4727 function formatSize( $size ) {
4728 return $this->formatComputingNumbers( $size, 1024, "size-$1bytes" );
4732 * Make a list item, used by various special pages
4734 * @param string $page Page link
4735 * @param string $details HTML safe text between brackets
4736 * @param bool $oppositedm Add the direction mark opposite to your
4737 * language, to display text properly
4738 * @return HTML escaped string
4740 function specialList( $page, $details, $oppositedm = true ) {
4745 $dirmark = ( $oppositedm ?
$this->getDirMark( true ) : '' ) . $this->getDirMark();
4749 $this->msg( 'word-separator' )->escaped() .
4750 $this->msg( 'parentheses' )->rawParams( $details )->escaped();
4754 * Generate (prev x| next x) (20|50|100...) type links for paging
4756 * @param Title $title Title object to link
4757 * @param int $offset
4759 * @param array $query Optional URL query parameter string
4760 * @param bool $atend Optional param for specified if this is the last page
4763 public function viewPrevNext( Title
$title, $offset, $limit,
4764 array $query = array(), $atend = false
4766 // @todo FIXME: Why on earth this needs one message for the text and another one for tooltip?
4768 # Make 'previous' link
4769 $prev = wfMessage( 'prevn' )->inLanguage( $this )->title( $title )->numParams( $limit )->text();
4770 if ( $offset > 0 ) {
4771 $plink = $this->numLink( $title, max( $offset - $limit, 0 ), $limit,
4772 $query, $prev, 'prevn-title', 'mw-prevlink' );
4774 $plink = htmlspecialchars( $prev );
4778 $next = wfMessage( 'nextn' )->inLanguage( $this )->title( $title )->numParams( $limit )->text();
4780 $nlink = htmlspecialchars( $next );
4782 $nlink = $this->numLink( $title, $offset +
$limit, $limit,
4783 $query, $next, 'nextn-title', 'mw-nextlink' );
4786 # Make links to set number of items per page
4787 $numLinks = array();
4788 foreach ( array( 20, 50, 100, 250, 500 ) as $num ) {
4789 $numLinks[] = $this->numLink( $title, $offset, $num,
4790 $query, $this->formatNum( $num ), 'shown-title', 'mw-numlink' );
4793 return wfMessage( 'viewprevnext' )->inLanguage( $this )->title( $title
4794 )->rawParams( $plink, $nlink, $this->pipeList( $numLinks ) )->escaped();
4798 * Helper function for viewPrevNext() that generates links
4800 * @param Title $title Title object to link
4801 * @param int $offset
4803 * @param array $query Extra query parameters
4804 * @param string $link Text to use for the link; will be escaped
4805 * @param string $tooltipMsg Name of the message to use as tooltip
4806 * @param string $class Value of the "class" attribute of the link
4807 * @return string HTML fragment
4809 private function numLink( Title
$title, $offset, $limit, array $query, $link,
4812 $query = array( 'limit' => $limit, 'offset' => $offset ) +
$query;
4813 $tooltip = wfMessage( $tooltipMsg )->inLanguage( $this )->title( $title )
4814 ->numParams( $limit )->text();
4816 return Html
::element( 'a', array( 'href' => $title->getLocalURL( $query ),
4817 'title' => $tooltip, 'class' => $class ), $link );
4821 * Get the conversion rule title, if any.
4825 public function getConvRuleTitle() {
4826 return $this->mConverter
->getConvRuleTitle();
4830 * Get the compiled plural rules for the language
4832 * @return array Associative array with plural form, and plural rule as key-value pairs
4834 public function getCompiledPluralRules() {
4835 $pluralRules = self
::$dataCache->getItem( strtolower( $this->mCode
), 'compiledPluralRules' );
4836 $fallbacks = Language
::getFallbacksFor( $this->mCode
);
4837 if ( !$pluralRules ) {
4838 foreach ( $fallbacks as $fallbackCode ) {
4839 $pluralRules = self
::$dataCache->getItem( strtolower( $fallbackCode ), 'compiledPluralRules' );
4840 if ( $pluralRules ) {
4845 return $pluralRules;
4849 * Get the plural rules for the language
4851 * @return array Associative array with plural form number and plural rule as key-value pairs
4853 public function getPluralRules() {
4854 $pluralRules = self
::$dataCache->getItem( strtolower( $this->mCode
), 'pluralRules' );
4855 $fallbacks = Language
::getFallbacksFor( $this->mCode
);
4856 if ( !$pluralRules ) {
4857 foreach ( $fallbacks as $fallbackCode ) {
4858 $pluralRules = self
::$dataCache->getItem( strtolower( $fallbackCode ), 'pluralRules' );
4859 if ( $pluralRules ) {
4864 return $pluralRules;
4868 * Get the plural rule types for the language
4870 * @return array Associative array with plural form number and plural rule type as key-value pairs
4872 public function getPluralRuleTypes() {
4873 $pluralRuleTypes = self
::$dataCache->getItem( strtolower( $this->mCode
), 'pluralRuleTypes' );
4874 $fallbacks = Language
::getFallbacksFor( $this->mCode
);
4875 if ( !$pluralRuleTypes ) {
4876 foreach ( $fallbacks as $fallbackCode ) {
4877 $pluralRuleTypes = self
::$dataCache->getItem( strtolower( $fallbackCode ), 'pluralRuleTypes' );
4878 if ( $pluralRuleTypes ) {
4883 return $pluralRuleTypes;
4887 * Find the index number of the plural rule appropriate for the given number
4888 * @param int $number
4889 * @return int The index number of the plural rule
4891 public function getPluralRuleIndexNumber( $number ) {
4892 $pluralRules = $this->getCompiledPluralRules();
4893 $form = CLDRPluralRuleEvaluator
::evaluateCompiled( $number, $pluralRules );
4898 * Find the plural rule type appropriate for the given number
4899 * For example, if the language is set to Arabic, getPluralType(5) should
4902 * @param int $number
4903 * @return string The name of the plural rule type, e.g. one, two, few, many
4905 public function getPluralRuleType( $number ) {
4906 $index = $this->getPluralRuleIndexNumber( $number );
4907 $pluralRuleTypes = $this->getPluralRuleTypes();
4908 if ( isset( $pluralRuleTypes[$index] ) ) {
4909 return $pluralRuleTypes[$index];